Search

Power distribution grids are typically installed outdoors and are exposed to environmental conditions. When contamination accumulates in the structures of the network, there may be shutdowns caused by electrical arcs. To improve the reliability of the network, visual inspections of the electrical power system can be carried out; these inspections can be automated using computer vision techniques based on deep neural networks. Based on this need, this paper proposes the Semi-ProtoPNet deep learning model to classify defective structures in the power distribution networks. The Semi-ProtoPNet deep neural network does not perform convex optimization of its last dense layer to maintain the impact of the negative reasoning process on image classification. The negative reasoning process rejects the incorrect classes of an input image; for this reason, it is possible to carry out an analysis with a low number of images that have different backgrounds, which is one of the challenges of this type of analysis. Semi-ProtoPNet achieves an accuracy of 97.22%, being superior to VGG-13, VGG-16, VGG-19, ResNet-34, ResNet-50, ResNet-152, DenseNet-121, DenseNet-161, DenseNet-201, and also models of the same class such as ProtoPNet, NP-ProtoPNet, Gen-ProtoPNet, and Ps-ProtoPNet. [ABSTRACT FROM AUTHOR]

Building extraction is a popular topic in remote sensing image processing. Efficient building extraction algorithms can identify and segment building areas to provide informative data for downstream tasks. Currently, building extraction is mainly achieved by deep convolutional neural networks (CNNs) based on the U-shaped encoder–decoder architecture. However, the local perceptive field of the convolutional operation poses a challenge for CNNs to fully capture the semantic information of large buildings, especially in high-resolution remote sensing images. Considering the recent success of the Transformer in computer vision tasks, in this paper, first we propose a shifted-window (swin) Transformer-based encoding booster. The proposed encoding booster includes a swin Transformer pyramid containing patch merging layers for down-sampling, which enables our encoding booster to extract semantics from multi-level features at different scales. Most importantly, the receptive field is significantly expanded by the global self-attention mechanism of the swin Transformer, allowing the encoding booster to capture the large-scale semantic information effectively and transcend the limitations of CNNs. Furthermore, we integrate the encoding booster in a specially designed U-shaped network through a novel manner, named the Swin Transformer-based Encoding Booster- U-shaped Network (STEB-UNet), to achieve the feature-level fusion of local and large-scale semantics. Remarkably, compared with other Transformer-included networks, the computational complexity and memory requirement of the STEB-UNet are significantly reduced due to the swin design, making the network training much easier. Experimental results show that the STEB-UNet can effectively discriminate and extract buildings of different scales and demonstrate higher accuracy than the state-of-the-art networks on public datasets. [ABSTRACT FROM AUTHOR]

Age estimation from human faces is an important yet challenging task in computer vision because of the large differences between physical age and apparent age. Due to the differences including races, genders, and other factors, the performance of a learning method for this task strongly depends on the training data. Although many inspiring works have focused on the age estimation of a single human face through deep learning, the existing methods still have lower performance when dealing with faces in videos because of the differences in head pose between frames, which can lead to greatly different results. In this paper, a combined system of age estimation and head pose estimation is proposed to improve the performance of age estimation from faces in videos. We use deep regression forests (DRFs) to estimate the age of facial images, while a multiloss convolutional neural network is also utilized to estimate the head pose. Accordingly, we estimate the age of faces only for head poses within a set degree threshold to enable value refinement. First, we divided the images in the Cross-Age Celebrity Dataset (CACD) and the Asian Face Age Dataset (AFAD) according to the estimated head pose degrees and generated separate age estimates for images with different poses. The experimental results showed that the accuracy of age estimation from frontal facial images was better than that for faces at different angles, thus demonstrating the effect of head pose on age estimation. Further experiments were conducted on several videos to estimate the age of the same person with his or her face at different angles, and the results show that our proposed combined system can provide more precise and reliable age estimates than a system without head pose estimation. [ABSTRACT FROM AUTHOR]

Computer vision has great potential to accelerate the global scale of photovoltaic potential analysis by extracting detailed roof information from high-resolution aerial images, but the lack of existing deep learning datasets is a major barrier. Therefore, we present the Roof Information Dataset for semantic segmentation of roof segments and roof superstructures. We assessed the label quality of initial roof superstructure annotations by conducting an annotation experiment and identified annotator agreements of 0.15–0.70 mean intersection over union, depending on the class. We discuss associated the implications on the training and evaluation of two convolutional neural networks and found that the quality of the prediction behaved similarly to the annotator agreement for most classes. The class photovoltaic module was predicted to be best with a class-specific mean intersection over union of 0.69. By providing the datasets in initial and reviewed versions, we promote a data-centric approach for the semantic segmentation of roof information. Finally, we conducted a photovoltaic potential analysis case study and demonstrated the high impact of roof superstructures as well as the viability of the computer vision approach to increase accuracy. While this paper's primary use case was roof information extraction for photovoltaic potential analysis, its implications can be transferred to other computer vision applications in remote sensing and beyond. [ABSTRACT FROM AUTHOR]

In this paper, we propose an activity detection system using a 24 × 32 resolution infrared array sensor placed on the ceiling. We first collect the data at different resolutions (i.e., 24 × 32, 12 × 16, and 6 × 8) and apply the advanced deep learning (DL) techniques of Super-Resolution (SR) and denoising to enhance the quality of the images. We then classify the images/sequences of images depending on the activities the subject is performing using a hybrid deep learning model combining a Convolutional Neural Network (CNN) and a Long Short-Term Memory (LSTM). We use data augmentation to improve the training of the neural networks by incorporating a wider variety of samples. The process of data augmentation is performed by a Conditional Generative Adversarial Network (CGAN). By enhancing the images using SR, removing the noise, and adding more training samples via data augmentation, our target is to improve the classification accuracy of the neural network. Through experiments, we show that employing these deep learning techniques to low-resolution noisy infrared images leads to a noticeable improvement in performance. The classification accuracy improved from 78.32% to 84.43% (for images with 6 × 8 resolution), and from 90.11% to 94.54% (for images with 12 × 16 resolution) when we used the CNN and CNN + LSTM networks, respectively. [ABSTRACT FROM AUTHOR]

In large coal preparation plants with a capacity of 30 million tons/year, the belt speed can reach 7 m/s and the thickness of the material layer can reach 500 mm. Therefore, in high-throughput and complex environments, the problem exists that harmful feeding materials such as iron and gangue are not easily detected, and thus fault diagnosis in the crushers lags behind. Therefore, it is necessary to extract the equipment operation signals from the noisy production environment and identify the feeding materials. Currently, there is no systematic research on signal processing and image classification of crusher feeding materials, while the convolutional neural network (CNN) is outstanding in computer vision. In this paper, sound and vibration signals of the feeding materials are denoised by spectral subtraction and transformed into feature images by continuous wavelet transforms. Then, an image classification model based on CNN is built for these feature images to study its classification mechanism and performance. The results show that the model classification accuracy is respectively 84.0%, 93.5% and 80.1% in coal–iron–wood classification, coal–iron classification, and coal–wood classification. The good classification performance for coal, iron and wood can satisfy the practical demands to remove the harmful feeding materials, which provides the core technical support for the establishment of operating status monitoring and fault diagnosis system of crushing equipment. [ABSTRACT FROM AUTHOR]

Facial emotion recognition (FER) is an important topic in the fields of computer vision and artificial intelligence owing to its significant academic and commercial potential. Although FER can be conducted using multiple sensors, this review focuses on studies that exclusively use facial images, because visual expressions are one of the main information channels in interpersonal communication. This paper provides a brief review of researches in the field of FER conducted over the past decades. First, conventional FER approaches are described along with a summary of the representative categories of FER systems and their main algorithms. Deep-learning-based FER approaches using deep networks enabling “end-to-end” learning are then presented. This review also focuses on an up-to-date hybrid deep-learning approach combining a convolutional neural network (CNN) for the spatial features of an individual frame and long short-term memory (LSTM) for temporal features of consecutive frames. In the later part of this paper, a brief review of publicly available evaluation metrics is given, and a comparison with benchmark results, which are a standard for a quantitative comparison of FER researches, is described. This review can serve as a brief guidebook to newcomers in the field of FER, providing basic knowledge and a general understanding of the latest state-of-the-art studies, as well as to experienced researchers looking for productive directions for future work. [ABSTRACT FROM AUTHOR]

Roads make a huge contribution to the economy and act as a platform for transportation. Potholes in roads are one of the major concerns in transportation infrastructure. A lot of research has proposed using computer vision techniques to automate pothole detection that include a wide range of image processing and object detection algorithms. There is a need to automate the pothole detection process with adequate accuracy and speed and implement the process easily and with low setup cost. In this paper, we have developed efficient deep learning convolution neural networks (CNNs) to detect potholes in real-time with adequate accuracy. To reduce the computational cost and improve the training results, this paper proposes a modified VGG16 (MVGG16) network by removing some convolution layers and using different dilation rates. Moreover, this paper uses the MVGG16 as a backbone network for the Faster R-CNN. In addition, this work compares the performance of YOLOv5 (Large (Yl), Medium (Ym), and Small (Ys)) models with ResNet101 backbone and Faster R-CNN with ResNet50(FPN), VGG16, MobileNetV2, InceptionV3, and MVGG16 backbones. The experimental results show that the Ys model is more applicable for real-time pothole detection because of its speed. In addition, using the MVGG16 network as the backbone of the Faster R-CNN provides better mean precision and shorter inference time than using VGG16, InceptionV3, or MobilNetV2 backbones. The proposed MVGG16 succeeds in balancing the pothole detection accuracy and speed. [ABSTRACT FROM AUTHOR]

Spaceborne synthetic aperture radar (SAR) represents a powerful source of data for enhancing maritime domain awareness (MDA). Wakes generated by traveling vessels hold a crucial role in MDA since they can be exploited both for ship route and velocity estimation and as a marker of ship presence. Even if deep learning (DL) has led to an impressive performance boost on a variety of computer vision tasks, its usage for automatic target recognition (ATR) in SAR images to support MDA is still limited to the detection of ships rather than ship wakes. A dataset is presented in this paper and several state-of-the-art object detectors based on convolutional neural networks (CNNs) are tested with different backbones. The dataset, including more than 250 wake chips, is realized by visually inspecting Sentinel-1 images over highly trafficked maritime sites. Extensive experiments are shown to characterize CNNs for the wake detection task. For the first time, a deep-learning approach is implemented to specifically detect ship wakes without any a-priori knowledge or cuing about the location of the vessel that generated the wake. No annotated dataset was available to train deep-learning detectors on this task, which is instead presented in this paper. Moreover, the benchmarks achieved for different detectors point out promising features and weak points of the relevant approaches. Thus, the work also aims at stimulating more research in this promising, but still under-investigated, field. [ABSTRACT FROM AUTHOR]

This paper's core objective is to develop and validate a new neurocomputing model to classify document images in particularly demanding hard conditions such as image distortions, image size variance and scale, a huge number of classes, etc. Document classification is a special machine vision task in which document images are categorized according to their likelihood. Document classification is by itself an important topic for the digital office and it has several usages. Additionally, different methods for solving this problem have been presented in various studies; their respectively reached performance is however not yet good enough. This task is very tough and challenging. Thus, a novel, more accurate and precise model is needed. Although the related works do reach acceptable accuracy values for less hard conditions, they generally fully fail in the face of those above-mentioned hard, real-world conditions, including, amongst others, distortions such as noise, blur, low contrast, and shadows. In this paper, a novel deep CNN model is developed, validated and benchmarked with a selection of the most relevant recent document classification models. Additionally, the model's sensitivity was significantly improved by injecting different artifacts during the training process. In the benchmarking, it does clearly outperform all others by at least 4%, thus reaching more than 96% accuracy. [ABSTRACT FROM AUTHOR]

Public aquariums and similar institutions often use video as a method to monitor the behavior, health, and status of aquatic organisms in their environments. These video footages take up a sizeable amount of space and require the use of autoencoders to reduce their file size for efficient storage. The autoencoder neural network is an emerging technique which uses the extracted latent space from an input source to reduce the image size for storage, and then reconstructs the source within an acceptable loss range for use. To meet an aquarium's practical needs, the autoencoder must have easily maintainable codes, low power consumption, be easily adoptable, and not require a substantial amount of memory use or processing power. Conventional configurations of autoencoders often provide results that perform beyond an aquarium's needs at the cost of being too complex for their architecture to handle, while few take low-contrast sources into consideration. Thus, in this instance, "keeping it simple" would be the ideal approach to the autoencoder's model design. This paper proposes a practical approach catered to an aquarium's specific needs through the configuration of autoencoder parameters. It first explores the differences between the two of the most widely applied autoencoder approaches, Multilayer Perceptron (MLP) and Convolution Neural Networks (CNN), to identify the most appropriate approach. The paper concludes that while both approaches (with proper configurations and image preprocessing) can reduce the dimensionality and reduce visual noise of the low-contrast images gathered from aquatic video footage, the CNN approach is more suitable for an aquarium's architecture. As an unexpected finding of the experiments conducted, the paper also discovered that by manipulating the formula for the MLP approach, the autoencoder could generate a denoised differential image that contains sharper and more desirable visual information to an aquarium's operation. Lastly, the paper has found that proper image preprocessing prior to the application of the autoencoder led to better model convergence and prediction results, as demonstrated both visually and numerically in the experiment. The paper concludes that by combining the denoising effect of MLP, CNN's ability to manage memory consumption, and proper image preprocessing, the specific practical needs of an aquarium can be adeptly fulfilled. [ABSTRACT FROM AUTHOR]

Insect monitoring methods are typically very time-consuming and involve substantial investment in species identification following manual trapping in the field. Insect traps are often only serviced weekly, resulting in low temporal resolution of the monitoring data, which hampers the ecological interpretation. This paper presents a portable computer vision system capable of attracting and detecting live insects. More specifically, the paper proposes detection and classification of species by recording images of live individuals attracted to a light trap. An Automated Moth Trap (AMT) with multiple light sources and a camera was designed to attract and monitor live insects during twilight and night hours. A computer vision algorithm referred to as Moth Classification and Counting (MCC), based on deep learning analysis of the captured images, tracked and counted the number of insects and identified moth species. Observations over 48 nights resulted in the capture of more than 250,000 images with an average of 5675 images per night. A customized convolutional neural network was trained on 2000 labeled images of live moths represented by eight different classes, achieving a high validation F 1 -score of 0.93. The algorithm measured an average classification and tracking F 1 -score of 0.71 and a tracking detection rate of 0.79. Overall, the proposed computer vision system and algorithm showed promising results as a low-cost solution for non-destructive and automatic monitoring of moths. [ABSTRACT FROM AUTHOR]

To reduce the loss induced by forest fires, it is very important to detect the forest fire smoke in real time so that early and timely warning can be issued. Machine vision and image processing technology is widely used for detecting forest fire smoke. However, most of the traditional image detection algorithms require manual extraction of image features and, thus, are not real-time. This paper evaluates the effectiveness of using the deep convolutional neural network to detect forest fire smoke in real time. Several target detection deep convolutional neural network algorithms evaluated include the EfficientDet (EfficientDet: Scalable and Efficient Object Detection), Faster R-CNN (Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks), YOLOv3 (You Only Look Once V3), and SSD (Single Shot MultiBox Detector) advanced CNN (Convolutional Neural Networks) model. The YOLOv3 showed a detection speed up to 27 FPS, indicating it is a real-time smoke detector. By comparing these algorithms with the current existing forest fire smoke detection algorithms, it can be found that the deep convolutional neural network algorithms result in better smoke detection accuracy. In particular, the EfficientDet algorithm achieves an average detection accuracy of 95.7%, which is the best real-time forest fire smoke detection among the evaluated algorithms. [ABSTRACT FROM AUTHOR]

The intelligent crack detection method is an important guarantee for the realization of intelligent operation and maintenance, and it is of great significance to traffic safety. In recent years, the recognition of road pavement cracks based on computer vision has attracted increasing attention. With the technological breakthroughs of general deep learning algorithms in recent years, detection algorithms based on deep learning and convolutional neural networks have achieved better results in the field of crack recognition. In this paper, deep learning is investigated to intelligently detect road cracks, and Faster R-CNN and Mask R-CNN are compared and analyzed. The results show that the joint training strategy is very effective, and we are able to ensure that both Faster R-CNN and Mask R-CNN complete the crack detection task when trained with only 130+ images and can outperform YOLOv3. However, the joint training strategy causes a degradation in the effectiveness of the bounding box detected by Mask R-CNN. [ABSTRACT FROM AUTHOR]

Illegal landfills are uncontrolled disposals of waste that cause severe environmental and health risk. Discovering them as early as possible is of prominent importance for preventing hazards, such as fire pollution and leakage. Before the digital era, the only means to detect illegal waste dumps was the on site inspection of potentially suspicious sites, a procedure extremely costly and impossible to scale to a vast territory. With the advent of Earth observation technology, scanning the territory via aerial images has become possible. However, manual image interpretation remains a complex and time-consuming task that requires expert skill. Photo interpretation can be partially automated by embedding the expert knowledge within a data driven classifier trained with samples provided by human annotators. In this paper, the detection of illegal landfills is formulated as a multi-scale scene classification problem. Scene elements positioning and spatial relations constitute hints of the presence of illegal waste dumps. A dataset of ≈3000 images (20 cm resolution per pixel) was created with the help of expert photo interpreters. A combination of ResNet50 and Feature Pyramid Network (FPN) elements accounting for different object scales achieves 88% precision with an 87% of recall in a test area. The results proved the feasibility of applying convolutional neural networks for scene classification in this scenario to optimize the process of waste dumps detection. [ABSTRACT FROM AUTHOR]

Underwater video surveys play a significant role in marine benthic research. Usually, surveys are filmed in transects, which are stitched into 2D mosaic maps for further analysis. Due to the massive amount of video data and time-consuming analysis, the need for automatic image segmentation and quantitative evaluation arises. This paper investigates such techniques on annotated mosaic maps containing hundreds of instances of brittle stars. By harnessing a deep convolutional neural network with pre-trained weights and post-processing results with a common blob detection technique, we investigate the effectiveness and potential of such segment-and-count approach by assessing the segmentation and counting success. Discs could be recommended instead of full shape masks for brittle stars due to faster annotation among marker variants tested. Underwater image enhancement techniques could not improve segmentation results noticeably, but some might be useful for augmentation purposes. [ABSTRACT FROM AUTHOR]

Recently, artificial intelligence has been successfully used in fields, such as computer vision, voice, and big data analysis. However, various problems, such as security, privacy, and ethics, also occur owing to the development of artificial intelligence. One such problem are deepfakes. Deepfake is a compound word for deep learning and fake. It refers to a fake video created using artificial intelligence technology or the production process itself. Deepfakes can be exploited for political abuse, pornography, and fake information. This paper proposes a method to determine integrity by analyzing the computer vision features of digital content. The proposed method extracts the rate of change in the computer vision features of adjacent frames and then checks whether the video is manipulated. The test demonstrated the highest detection rate of 97% compared to the existing method or machine learning method. It also maintained the highest detection rate of 96%, even for the test that manipulates the matrix of the image to avoid the convolutional neural network detection method. [ABSTRACT FROM AUTHOR]

Artificial neural networks have become the go-to solution for computer vision tasks, including problems of the security domain. One such example comes in the form of reidentification, where deep learning can be part of the surveillance pipeline. The use case necessitates considering an adversarial setting—and neural networks have been shown to be vulnerable to a range of attacks. In this paper, the preprocessing defences against adversarial attacks are evaluated, including block-matching convolutional neural network for image denoising used as an adversarial defence. The benefit of using preprocessing defences comes from the fact that it does not require the effort of retraining the classifier, which, in computer vision problems, is a computationally heavy task. The defences are tested in a real-life-like scenario of using a pre-trained, widely available neural network architecture adapted to a specific task with the use of transfer learning. Multiple preprocessing pipelines are tested and the results are promising. [ABSTRACT FROM AUTHOR]

In turning, the wear control of a cutting tool benefits product quality enhancement, tool-related costs' optimisation, and assists in avoiding undesired events. In small series and individual production, the machine operator is the one who determines when to change a cutting tool, based upon their experience. Bad decisions can often lead to greater costs, production downtime, and scrap. In this paper, a Tool Condition Monitoring (TCM) system is presented that automatically classifies tool wear of turning tools into four classes (no, low, medium, high wear). A cutting tool was monitored with infrared (IR) camera immediately after the cut and in the following 60 s. The Convolutional Neural Network Inception V3 was used to analyse and classify the thermographic images, which were divided into different groups depending on the time of acquisition. Based on classification result, one gets information about the cutting capability of the tool for further machining. The proposed model, combining Infrared Thermography, Computer Vision, and Deep Learning, proved to be a suitable method with results of more than 96% accuracy. The most appropriate time of image acquisition is 6–12 s after the cut is finished. While existing temperature based TCM systems focus on measuring a cutting tool absolute temperature, the proposed system analyses a temperature distribution (relative temperatures) on the whole image based on image features. [ABSTRACT FROM AUTHOR]

Automated pavement distress recognition is a key step in smart infrastructure assessment. Advances in deep learning and computer vision have improved the automated recognition of pavement distresses in road surface images. This task remains challenging due to the high variation of defects in shapes and sizes, demanding a better incorporation of contextual information into deep networks. In this paper, we show that an attention-based multi-scale convolutional neural network (A+MCNN) improves the automated classification of common distress and non-distress objects in pavement images by (i) encoding contextual information through multi-scale input tiles and (ii) employing a mid-fusion approach with an attention module for heterogeneous image contexts from different input scales. A+MCNN is trained and tested with four distress classes (crack, crack seal, patch, pothole), five non-distress classes (joint, marker, manhole cover, curbing, shoulder), and two pavement classes (asphalt, concrete). A+MCNN is compared with four deep classifiers that are widely used in transportation applications and a generic CNN classifier (as the control model). The results show that A+MCNN consistently outperforms the baselines by 1∼26% on average in terms of the F-score. A comprehensive discussion is also presented regarding how these classifiers perform differently on different road objects, which has been rarely addressed in the existing literature. [ABSTRACT FROM AUTHOR]

The automation of lifespan assays with C. elegans in standard Petri dishes is a challenging problem because there are several problems hindering detection such as occlusions at the plate edges, dirt accumulation, and worm aggregations. Moreover, determining whether a worm is alive or dead can be complex as they barely move during the last few days of their lives. This paper proposes a method combining traditional computer vision techniques with a live/dead C. elegans classifier based on convolutional and recurrent neural networks from low-resolution image sequences. In addition to proposing a new method to automate lifespan, the use of data augmentation techniques is proposed to train the network in the absence of large numbers of samples. The proposed method achieved small error rates (3.54% ± 1.30% per plate) with respect to the manual curve, demonstrating its feasibility. [ABSTRACT FROM AUTHOR]

Drilling operations are an essential part of furniture from MDF laminated boards required for product assembly. Faults in the process might introduce adverse effects to the furniture. Inspection of the drilling quality can be challenging due to a big variety of board surface textures, dust, or woodchips in the manufacturing process, milling cutouts, and other kinds of defects. Intelligent computer vision methods can be engaged for global contextual analysis with local information attention for automated object detection and segmentation. In this paper, we propose blind and through drilled holes segmentation on textured wooden furniture panel images using the UNet encoder-decoder modifications enhanced with residual connections, atrous spatial pyramid pooling, squeeze and excitation module, and CoordConv layers for better segmentation performance. We show that even a lightweight architecture is capable to perform on a range of complex textures and is able to distinguish the holes drilling operations' semantical information from the rest of the furniture board and conveyor context. The proposed model configurations yield better results in more complex cases with a not significant or small bump in processing time. Experimental results demonstrate that our best-proposed solution achieves a Dice score of up to 97.89% compared to the baseline U-Net model's Dice score of 94.50%. Statistical, visual, and computational properties of each convolutional neural network architecture are addressed. [ABSTRACT FROM AUTHOR]

The estimation of player positions is key for performance analysis in sport. In this paper, we focus on image-based, single-angle, player position estimation in padel. Unlike tennis, the primary camera view in professional padel videos follows a de facto standard, consisting of a high-angle shot at about 7.6 m above the court floor. This camera angle reduces the occlusion impact of the mesh that stands over the glass walls, and offers a convenient view for judging the depth of the ball and the player positions and poses. We evaluate and compare the accuracy of state-of-the-art computer vision methods on a large set of images from both amateur videos and publicly available videos from the major international padel circuit. The methods we analyze include object detection, image segmentation and pose estimation techniques, all of them based on deep convolutional neural networks. We report accuracy and average precision with respect to manually-annotated video frames. The best results are obtained by top-down pose estimation methods, which offer a detection rate of 99.8% and a RMSE below 5 and 12 cm for horizontal/vertical court-space coordinates (deviations from predicted and ground-truth player positions). These results demonstrate the suitability of pose estimation methods based on deep convolutional neural networks for estimating player positions from single-angle padel videos. Immediate applications of this work include the player and team analysis of the large collection of publicly available videos from international circuits, as well as an inexpensive method to get player positional data in amateur padel clubs. [ABSTRACT FROM AUTHOR]

Super resolution (SR) enables to generate a high-resolution (HR) image from one or more low-resolution (LR) images. Since a variety of CNN models have been recently studied in the areas of computer vision, these approaches have been combined with SR in order to provide higher image restoration. In this paper, we propose a lightweight CNN-based SR method, named multi-scale channel dense network (MCDN). In order to design the proposed network, we extracted the training images from the DIVerse 2K (DIV2K) dataset and investigated the trade-off between the SR accuracy and the network complexity. The experimental results show that the proposed method can significantly reduce the network complexity, such as the number of network parameters and total memory capacity, while maintaining slightly better or similar perceptual quality compared to the previous methods. [ABSTRACT FROM AUTHOR]

The new coronavirus disease (COVID-19), pneumonia, tuberculosis, and breast cancer have one thing in common: these diseases can be diagnosed using radiological studies such as X-rays images. With radiological studies and technology, computer-aided diagnosis (CAD) results in a very useful technique to analyze and detect abnormalities using the images generated by X-ray machines. Some deep-learning techniques such as a convolutional neural network (CNN) can help physicians to obtain an effective pre-diagnosis. However, popular CNNs are enormous models and need a huge amount of data to obtain good results. In this paper, we introduce NanoChest-net, which is a small but effective CNN model that can be used to classify among different diseases using images from radiological studies. NanoChest-net proves to be effective in classifying among different diseases such as tuberculosis, pneumonia, and COVID-19. In two of the five datasets used in the experiments, NanoChest-net obtained the best results, while on the remaining datasets our model proved to be as good as baseline models from the state of the art such as the ResNet50, Xception, and DenseNet121. In addition, NanoChest-net is useful to classify radiological studies on the same level as state-of-the-art algorithms with the advantage that it does not require a large number of operations. [ABSTRACT FROM AUTHOR]

A vital and challenging task in computer vision is 3D Object Classification and Retrieval, with many practical applications such as an intelligent robot, autonomous driving, multimedia contents processing and retrieval, and augmented/mixed reality. Various deep learning methods were introduced for solving classification and retrieval problems of 3D objects. Almost all view-based methods use many views to handle spatial loss, although they perform the best among current techniques such as View-based, Voxelization, and Point Cloud methods. Many views make network structure more complicated due to the parallel Convolutional Neural Network (CNN). We propose a novel method that combines a Global Point Signature Plus with a Deep Wide Residual Network, namely GPSP-DWRN, in this paper. Global Point Signature Plus (GPSPlus) is a novel descriptor because it can capture more shape information of the 3D object for a single view. First, an original 3D model was converted into a colored one by applying GPSPlus. Then, a 32 × 32 × 3 matrix stored the obtained 2D projection of this color 3D model. This matrix was the input data of a Deep Residual Network, which used a single CNN structure. We evaluated the GPSP-DWRN for a retrieval task using the Shapnetcore55 dataset, while using two well-known datasets—ModelNet10 and ModelNet40 for a classification task. Based on our experimental results, our framework performed better than the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In recent years, the technology of artificial intelligence (AI) and robots is rapidly spreading to countries around the world. More and more scholars and industry experts have proposed AI deep learning models and methods to solve human life problems and improve work efficiency. Modern people's lives are very busy, which led us to investigate whether the demand for Bento buffet cafeterias has gradually increased in Taiwan. However, when eating at a buffet in a cafeteria, people often encounter two problems. The first problem is that customers need to queue up to check out after they have selected and filled their dishes from the buffet. However, it always takes too much time waiting, especially at lunch or dinner time. The second problem is sometimes customers question the charges calculated by cafeteria staff, claiming they are too expensive at the checkout counter. Therefore, it is necessary to develop an AI-enabled checkout system. The AI-enabled self-checkout system will help the Bento buffet cafeterias reduce long lineups without the need to add additional workers. In this paper, we used computer vision and deep-learning technology to design and implement an AI-enabled checkout system for Bento buffet cafeterias. The prototype contains an angle steel shelf, a Kinect camera, a light source, and a desktop computer. Six baseline convolutional neural networks were applied for comparison on food recognition. In our experiments, there were 22 different food categories in a Bento buffet cafeteria employed. Experimental results show that the inception_v4 model can achieve the highest average validation accuracy of 99.11% on food recognition, but it requires the most training and recognition time. AlexNet model achieves a 94.5% accuracy and requires the least training time and recognition time. We propose a hierarchical approach with two stages to achieve good performance in both the recognition accuracy rate and the required training and recognition time. The approach is designed to perform the first step of identification and the second step of recognizing similar food images, respectively. Experimental results show that the proposed approach can achieve a 96.3% accuracy rate on our test dataset and required very little recognition time for input images. In addition, food volumes could be estimated using the depth images captured by the Kinect camera, and a framework of visual checkout system was successfully built. [ABSTRACT FROM AUTHOR]

Deep-learning methods, especially convolutional neural networks (CNN), have become the first choice for hyperspectral image (HSI) classification to date. It is a common procedure that small cubes are cropped from hyperspectral images and then fed into CNNs. However, standard CNNs find it difficult to extract discriminative spectral–spatial features. How to obtain finer spectral–spatial features to improve the classification performance is now a hot topic of research. In this regard, the attention mechanism, which has achieved excellent performance in other computer vision, holds the exciting prospect. In this paper, we propose a double-branch network consisting of a novel convolution named pyramidal convolution (PyConv) and an iterative attention mechanism. Each branch concentrates on exploiting spectral or spatial features with different PyConvs, supplemented by the attention module for refining the feature map. Experimental results demonstrate that our model can yield competitive performance compared to other state-of-the-art models. [ABSTRACT FROM AUTHOR]

In recent years, benefiting from the rapid development of deep learning technology in the field of computer vision, the study of hyperspectral image (HSI) classification has also made great progress. However, compared with ordinary RGB images, HSIs are more like 3D cubes; therefore, it is necessary and beneficial to explore classification methods suitable for the very special data structure of HSIs. In this paper, we propose Multiple Spectral Resolution 3D Convolutional Neural Network (MSR-3DCNN) for HSI classification tasks. In MSR-3DCNN, we expand the idea of multi-scale feature fusion and dilated convolution from the spatial dimension to the spectral dimension, and combine 3D convolution and residual connection; therefore, it can better adapt to the 3D cubic form of hyperspectral data and make efficient use of spectral information in different bands. Experimental results on four benchmark datasets show the effectiveness of the proposed approach and its superiority as compared with some state-of-the-art (SOTA) HSI classification methods. [ABSTRACT FROM AUTHOR]

Semantic segmentation, which refers to pixel-wise classification of an image, is a fundamental topic in computer vision owing to its growing importance in the robot vision and autonomous driving sectors. It provides rich information about objects in the scene such as object boundary, category, and location. Recent methods for semantic segmentation often employ an encoder-decoder structure using deep convolutional neural networks. The encoder part extracts features of the image using several filters and pooling operations, whereas the decoder part gradually recovers the low-resolution feature maps of the encoder into a full input resolution feature map for pixel-wise prediction. However, the encoder-decoder variants for semantic segmentation suffer from severe spatial information loss, caused by pooling operations or stepwise convolutions, and does not consider the context in the scene. In this paper, we propose a novel dense upsampling convolution method based on a guided filter to effectively preserve the spatial information of the image in the network. We further propose a novel local context convolution method that not only covers larger-scale objects in the scene but covers them densely for precise object boundary delineation. Theoretical analyses and experimental results on several benchmark datasets verify the effectiveness of our method. Qualitatively, our approach delineates object boundaries at a level of accuracy that is beyond the current excellent methods. Quantitatively, we report a new record of 82.86% and 81.62% of pixel accuracy on ADE20K and Pascal-Context benchmark datasets, respectively. In comparison with the state-of-the-art methods, the proposed method offers promising improvements. [ABSTRACT FROM AUTHOR]

Hyperspectral image (HSI) classification is the subject of intense research in remote sensing. The tremendous success of deep learning in computer vision has recently sparked the interest in applying deep learning in hyperspectral image classification. However, most deep learning methods for hyperspectral image classification are based on convolutional neural networks (CNN). Those methods require heavy GPU memory resources and run time. Recently, another deep learning model, the transformer, has been applied for image recognition, and the study result demonstrates the great potential of the transformer network for computer vision tasks. In this paper, we propose a model for hyperspectral image classification based on the transformer, which is widely used in natural language processing. Besides, we believe we are the first to combine the metric learning and the transformer model in hyperspectral image classification. Moreover, to improve the model classification performance when the available training samples are limited, we use the 1-D convolution and Mish activation function. The experimental results on three widely used hyperspectral image data sets demonstrate the proposed model's advantages in accuracy, GPU memory cost, and running time. [ABSTRACT FROM AUTHOR]

Deep learning methods have been successfully applied to image processing, mainly using 2D vision sensors. Recently, the rise of depth cameras and other similar 3D sensors has opened the field for new perception techniques. Nevertheless, 3D convolutional neural networks perform slightly worse than other 3D deep learning methods, and even worse than their 2D version. In this paper, we propose to improve 3D deep learning results by transferring the pretrained weights learned in 2D networks to their corresponding 3D version. Using an industrial object recognition context, we have analyzed different combinations of 3D convolutional networks (VGG16, ResNet, Inception ResNet, and EfficientNet), comparing the recognition accuracy. The highest accuracy is obtained with EfficientNetB0 using extrusion with an accuracy of 0.9217, which gives comparable results to state-of-the art methods. We also observed that the transfer approach enabled to improve the accuracy of the Inception ResNet 3D version up to 18% with respect to the score of the 3D approach alone. [ABSTRACT FROM AUTHOR]

The potential benefits of recognising activities of daily living from video for active and assisted living have yet to be fully untapped. These technologies can be used for behaviour understanding, and lifelogging for caregivers and end users alike. The recent publication of realistic datasets for this purpose, such as the Toyota Smarthomes dataset, calls for pushing forward the efforts to improve action recognition. Using the separable spatio-temporal attention network proposed in the literature, this paper introduces a view-invariant normalisation of skeletal pose data and full activity crops for RGB data, which improve the baseline results by 9.5% (on the cross-subject experiments), outperforming state-of-the-art techniques in this field when using the original unmodified skeletal data in dataset. Our code and data are available online. [ABSTRACT FROM AUTHOR]

Sand theft or illegal mining in river dredging areas has been a problem in recent decades. For this reason, increasing the use of artificial intelligence in dredging areas, building automated monitoring systems, and reducing human involvement can effectively deter crime and lighten the workload of security guards. In this investigation, a smart dredging construction site system was developed using automated techniques that were arranged to be suitable to various areas. The aim in the initial period of the smart dredging construction was to automate the audit work at the control point, which manages trucks in river dredging areas. Images of dump trucks entering the control point were captured using monitoring equipment in the construction area. The obtained images and the deep learning technique, YOLOv3, were used to detect the positions of the vehicle license plates. Framed images of the vehicle license plates were captured and were used as input in an image classification model, C-CNN-L3, to identify the number of characters on the license plate. Based on the classification results, the images of the vehicle license plates were transmitted to a text recognition model, R-CNN-L3, that corresponded to the characters of the license plate. Finally, the models of each stage were integrated into a real-time truck license plate recognition (TLPR) system; the single character recognition rate was 97.59%, the overall recognition rate was 93.73%, and the speed was 0.3271 s/image. The TLPR system reduces the labor force and time spent to identify the license plates, effectively reducing the probability of crime and increasing the transparency, automation, and efficiency of the frontline personnel's work. The TLPR is the first step toward an automated operation to manage trucks at the control point. The subsequent and ongoing development of system functions can advance dredging operations toward the goal of being a smart construction site. By intending to facilitate an intelligent and highly efficient management system of dredging-related departments by providing a vehicle LPR system, this paper forms a contribution to the current body of knowledge in the sense that it presents an objective approach for the TLPR system. [ABSTRACT FROM AUTHOR]

Due to deep learning's accurate cognition of the street environment, the convolutional neural network has achieved dramatic development in the application of street scenes. Considering the needs of autonomous driving and assisted driving, in a general way, computer vision technology is used to find obstacles to avoid collisions, which has made semantic segmentation a research priority in recent years. However, semantic segmentation has been constantly facing new challenges for quite a long time. Complex network depth information, large datasets, real-time requirements, etc., are typical problems that need to be solved urgently in the realization of autonomous driving technology. In order to address these problems, we propose an improved lightweight real-time semantic segmentation network, which is based on an efficient image cascading network (ICNet) architecture, using multi-scale branches and a cascaded feature fusion unit to extract rich multi-level features. In this paper, a spatial information network is designed to transmit more prior knowledge of spatial location and edge information. During the course of the training phase, we append an external loss function to enhance the learning process of the deep learning network system as well. This lightweight network can quickly perceive obstacles and detect roads in the drivable area from images to satisfy autonomous driving characteristics. The proposed model shows substantial performance on the Cityscapes dataset. With the premise of ensuring real-time performance, several sets of experimental comparisons illustrate that SP-ICNet enhances the accuracy of road obstacle detection and provides nearly ideal prediction outputs. Compared to the current popular semantic segmentation network, this study also demonstrates the effectiveness of our lightweight network for road obstacle detection in autonomous driving. [ABSTRACT FROM AUTHOR]

With the increasing popularity of artificial intelligence, deep learning has been applied to various fields, especially in computer vision. Since artificial intelligence is migrating from cloud to edge, deep learning nowadays should be edge-oriented and adaptive to complex environments. Aiming at these goals, this paper proposes an ICONet (illumination condition optimized network). Based on OTSU segmentation algorithm and fuzzy c-means clustering algorithm, the illumination condition classification subnet increases the environmental adaptivity of our network. The reduced time complexity and optimized size of our convolutional neural network (CNN) model enables the implementation of ICONet on edge devices. In the field of fatigue driving, we test the performance of ICONet on YawDD and self-collected datasets. Our network achieves a general accuracy of 98.56% and our models are about 590 kilobytes. Compared to other proposed networks, the ICONet shows significant success and superiority. Applying ICONet to fatigue driving detection is helpful to solve the symmetry of the needs of edge-oriented detection under complex illumination condition environments and the scarcity of related approaches. [ABSTRACT FROM AUTHOR]

Due to recent advancements in virtual reality (VR) and augmented reality (AR), the demand for high quality immersive contents is a primary concern for production companies and consumers. Similarly, the topical record-breaking performance of deep learning in various domains of artificial intelligence has extended the attention of researchers to contribute to different fields of computer vision. To ensure the quality of immersive media contents using these advanced deep learning technologies, several learning based Stitched Image Quality Assessment methods have been proposed with reasonable performances. However, these methods are unable to localize, segment, and extract the stitching errors in panoramic images. Further, these methods used computationally complex procedures for quality assessment of panoramic images. With these motivations, in this paper, we propose a novel three-fold Deep Learning based No-Reference Stitched Image Quality Assessment (DLNR-SIQA) approach to evaluate the quality of immersive contents. In the first fold, we fined-tuned the state-of-the-art Mask R-CNN (Regional Convolutional Neural Network) on manually annotated various stitching error-based cropped images from the two publicly available datasets. In the second fold, we segment and localize various stitching errors present in the immersive contents. Finally, based on the distorted regions present in the immersive contents, we measured the overall quality of the stitched images. Unlike existing methods that only measure the quality of the images using deep features, our proposed method can efficiently segment and localize stitching errors and estimate the image quality by investigating segmented regions. We also carried out extensive qualitative and quantitative comparison with full reference image quality assessment (FR-IQA) and no reference image quality assessment (NR-IQA) on two publicly available datasets, where the proposed system outperformed the existing state-of-the-art techniques. [ABSTRACT FROM AUTHOR]

Unmanned Aerial Vehicles (UAVs) have been very effective in collecting aerial images data for various Internet-of-Things (IoT)/smart cities applications such as search and rescue, surveillance, vehicle detection, counting, intelligent transportation systems, to name a few. However, the real-time processing of collected data on edge in the context of the Internet-of-Drones remains an open challenge because UAVs have limited energy capabilities, while computer vision techniquesconsume excessive energy and require abundant resources. This fact is even more critical when deep learning algorithms, such as convolutional neural networks (CNNs), are used for classification and detection. In this paper, we first propose a system architecture of computation offloading for Internet-connected drones. Then, we conduct a comprehensive experimental study to evaluate the performance in terms of energy, bandwidth, and delay of the cloud computation offloading approach versus the edge computing approach of deep learning applications in the context of UAVs. In particular, we investigate the tradeoff between the communication cost and the computation of the two candidate approaches experimentally. The main results demonstrate that the computation offloading approach allows us to provide much higher throughput (i.e., frames per second) as compared to the edge computing approach, despite the larger communication delays. [ABSTRACT FROM AUTHOR]

The recognition of human activities is usually considered to be a simple procedure. Problems occur in complex scenes involving high speeds. Activity prediction using Artificial Intelligence (AI) by numerical analysis has attracted the attention of several researchers. Human activities are an important challenge in various fields. There are many great applications in this area, including smart homes, assistive robotics, human–computer interactions, and improvements in protection in several areas such as security, transport, education, and medicine through the control of falling or aiding in medication consumption for elderly people. The advanced enhancement and success of deep learning techniques in various computer vision applications encourage the use of these methods in video processing. The human presentation is an important challenge in the analysis of human behavior through activity. A person in a video sequence can be described by their motion, skeleton, and/or spatial characteristics. In this paper, we present a novel approach to human activity recognition from videos using the Recurrent Neural Network (RNN) for activity classification and the Convolutional Neural Network (CNN) with a new structure of the human skeleton to carry out feature presentation. The aims of this work are to improve the human presentation through the collection of different features and the exploitation of the new RNN structure for activities. The performance of the proposed approach is evaluated by the RGB-D sensor dataset CAD-60. The experimental results show the performance of the proposed approach through the average error rate obtained (4.5%). [ABSTRACT FROM AUTHOR]

Land cover classification with the focus on chlorophyll-rich vegetation detection plays an important role in urban growth monitoring and planning, autonomous navigation, drone mapping, biodiversity conservation, etc. Conventional approaches usually apply the normalized difference vegetation index (NDVI) for vegetation detection. In this paper, we investigate the performance of deep learning and conventional methods for vegetation detection. Two deep learning methods, DeepLabV3+ and our customized convolutional neural network (CNN) were evaluated with respect to their detection performance when training and testing datasets originated from different geographical sites with different image resolutions. A novel object-based vegetation detection approach, which utilizes NDVI, computer vision, and machine learning (ML) techniques, is also proposed. The vegetation detection methods were applied to high-resolution airborne color images which consist of RGB and near-infrared (NIR) bands. RGB color images alone were also used with the two deep learning methods to examine their detection performances without the NIR band. The detection performances of the deep learning methods with respect to the object-based detection approach are discussed and sample images from the datasets are used for demonstrations. [ABSTRACT FROM AUTHOR]

In the manuscript, the issue of detecting and segmenting out pavement defects on highway roads is addressed. Specifically, computer vision (CV) methods are developed and applied to the problem based on deep learning of convolutional neural networks (ConvNets). A novel neural network structure is considered, based on a pipeline of three ConvNets and endowed with the capacity for context awareness, which improves grid-based search for defects on orthoframes by considering the surrounding image content—an approach, which essentially draws inspiration from how humans tend to solve the task of image segmentation. Also, methods for assessing the quality of segmentation are discussed. The contribution also describes the complete procedure of working with pavement defects in an industrial setting, involving the workcycle of defect annotation, ConvNet training and validation. The results of ConvNet evaluation provided in the paper hint at a successful implementation of the proposed technique. [ABSTRACT FROM AUTHOR]

With the increasing number of drones, the danger of their illegal use has become relevant. This has necessitated the creation of automatic drone protection systems. One of the important tasks solved by these systems is the reliable detection of drones near guarded objects. This problem can be solved using various methods. From the point of view of the price–quality ratio, the use of video cameras for a drone detection is of great interest. However, drone detection using visual information is hampered by the large similarity of drones to other objects, such as birds or airplanes. In addition, drones can reach very high speeds, so detection should be done in real time. This paper addresses the problem of real-time drone detection with high accuracy. We divided the drone detection task into two separate tasks: the detection of moving objects and the classification of the detected object into drone, bird, and background. The moving object detection is based on background subtraction, while classification is performed using a convolutional neural network (CNN). The experimental results showed that the proposed approach can achieve an accuracy comparable to existing approaches at high processing speed. We also concluded that the main limitation of our detector is the dependence of its performance on the presence of a moving background. [ABSTRACT FROM AUTHOR]

Determining an object measurement is a challenging task without having a well-defined reference. When a ruler is placed in the same plane of an object being measured it can serve as metric reference, thus a measurement system can be defined and calibrated to correlate actual dimensions with pixels contained in an image. This paper describes a system for non-contact object measurement by sensing and assessing the distinct spatial frequency of the graduations on a ruler. The approach presented leverages Deep Learning methods, specifically Mask Region proposal based Convolutional Neural Networks (R-CNN), for rulers' recognition and segmentation, as well as several other computer vision (CV) methods such as adaptive thresholding and template matching. We developed a heuristic analytical method for calibrating an image by applying several filters to extract the spatial frequencies corresponding to the ticks on a given ruler. We propose an automated in-plane optical scaling calibration system for non-contact measurement. [ABSTRACT FROM AUTHOR]

Facial expression recognition (FER) is a challenging problem in the fields of pattern recognition and computer vision. The recent success of convolutional neural networks (CNNs) in object detection and object segmentation tasks has shown promise in building an automatic deep CNN-based FER model. However, in real-world scenarios, performance degrades dramatically owing to the great diversity of factors unrelated to facial expressions, and due to a lack of training data and an intrinsic imbalance in the existing facial emotion datasets. To tackle these problems, this paper not only applies deep transfer learning techniques, but also proposes a novel loss function called weighted-cluster loss, which is used during the fine-tuning phase. Specifically, the weighted-cluster loss function simultaneously improves the intra-class compactness and the inter-class separability by learning a class center for each emotion class. It also takes the imbalance in a facial expression dataset into account by giving each emotion class a weight based on its proportion of the total number of images. In addition, a recent, successful deep CNN architecture, pre-trained in the task of face identification with the VGGFace2 database from the Visual Geometry Group at Oxford University, is employed and fine-tuned using the proposed loss function to recognize eight basic facial emotions from the AffectNet database of facial expression, valence, and arousal computing in the wild. Experiments on an AffectNet real-world facial dataset demonstrate that our method outperforms the baseline CNN models that use either weighted-softmax loss or center loss. [ABSTRACT FROM AUTHOR]

The interaction between humans and an NAO robot using deep convolutional neural networks (CNN) is presented in this paper based on an innovative end-to-end pipeline method that applies two optimized CNNs, one for face recognition (FR) and another one for the facial expression recognition (FER) in order to obtain real-time inference speed for the entire process. Two different models for FR are considered, one known to be very accurate, but has low inference speed (faster region-based convolutional neural network), and one that is not as accurate but has high inference speed (single shot detector convolutional neural network). For emotion recognition transfer learning and fine-tuning of three CNN models (VGG, Inception V3 and ResNet) has been used. The overall results show that single shot detector convolutional neural network (SSD CNN) and faster region-based convolutional neural network (Faster R-CNN) models for face detection share almost the same accuracy: 97.8% for Faster R-CNN on PASCAL visual object classes (PASCAL VOCs) evaluation metrics and 97.42% for SSD Inception. In terms of FER, ResNet obtained the highest training accuracy (90.14%), while the visual geometry group (VGG) network had 87% accuracy and Inception V3 reached 81%. The results show improvements over 10% when using two serialized CNN, instead of using only the FER CNN, while the recent optimization model, called rectified adaptive moment optimization (RAdam), lead to a better generalization and accuracy improvement of 3%-4% on each emotion recognition CNN. [ABSTRACT FROM AUTHOR]

Deep learning has already been proved as a powerful state-of-the-art technique for many image understanding tasks in computer vision and other applications including remote sensing (RS) image analysis. Unmanned aircraft systems (UASs) offer a viable and economical alternative to a conventional sensor and platform for acquiring high spatial and high temporal resolution data with high operational flexibility. Coastal wetlands are among some of the most challenging and complex ecosystems for land cover prediction and mapping tasks because land cover targets often show high intra-class and low inter-class variances. In recent years, several deep convolutional neural network (CNN) architectures have been proposed for pixel-wise image labeling, commonly called semantic image segmentation. In this paper, some of the more recent deep CNN architectures proposed for semantic image segmentation are reviewed, and each model's training efficiency and classification performance are evaluated by training it on a limited labeled image set. Training samples are provided using the hyper-spatial resolution UAS imagery over a wetland area and the required ground truth images are prepared by manual image labeling. Experimental results demonstrate that deep CNNs have a great potential for accurate land cover prediction task using UAS hyper-spatial resolution images. Some simple deep learning architectures perform comparable or even better than complex and very deep architectures with remarkably fewer training epochs. This performance is especially valuable when limited training samples are available, which is a common case in most RS applications. [ABSTRACT FROM AUTHOR]

With the rapid development of surveying and spatial information technologies, more and more attention has been given to positioning. In outdoor environments, people can easily obtain positioning services through global navigation satellite systems (GNSS). In indoor environments, the GNSS signal is often lost, while other positioning problems, such as dead reckoning and wireless signals, will face accumulated errors and signal interference. Therefore, this research uses images to realize a positioning service. The main concept of this work is to establish a model for an indoor field image and its coordinate information and to judge its position by image eigenvalue matching. Based on the architecture of PoseNet, the image is input into a 23-layer convolutional neural network according to various sizes to train end-to-end location identification tasks, and the three-dimensional position vector of the camera is regressed. The experimental data are taken from the underground parking lot and the Palace Museum. The preliminary experimental results show that this new method designed by us can effectively improve the accuracy of indoor positioning by about 20% to 30%. In addition, this paper also discusses other architectures, field sizes, camera parameters, and error corrections for this neural network system. The preliminary experimental results show that the angle error correction method designed by us can effectively improve positioning by about 20%. [ABSTRACT FROM AUTHOR]

Smoke detection technology based on computer vision is a popular research direction in fire detection. This technology is widely used in outdoor fire detection fields (e.g., forest fire detection). Smoke detection is often based on features such as color, shape, texture, and motion to distinguish between smoke and non-smoke objects. However, the salience and robustness of these features are insufficiently strong, resulting in low smoke detection performance under complex environment. Deep learning technology has improved smoke detection performance to a certain degree, but extracting smoke detail features is difficult when the number of network layers is small. With no effective use of smoke motion characteristics, indicators such as false alarm rate are high in video smoke detection. To enhance the detection performance of smoke objects in videos, this paper proposes a concept of change-cumulative image by converting the YUV color space of multi-frame video images into a change-cumulative image, which can represent the motion and color-change characteristics of smoke. Then, a fusion deep network is designed, which increases the depth of the VGG16 network by arranging two convolutional layers after each of its convolutional layer. The VGG16 and Resnet50 (Deep residual network) network models are also arranged using the fusion deep network to improve feature expression ability while increasing the depth of the whole network. Doing so can help extract additional discriminating characteristics of smoke. Experimental results show that by using the change-cumulative image as the input image of the deep network model, smoke detection performance is superior to the classic RGB input image; the smoke detection performance of the fusion deep network model is better than that of the single VGG16 and Resnet50 network models; the smoke detection accuracy, false positive rate, and false alarm rate of this method are better than those of the current popular methods of video smoke detection. [ABSTRACT FROM AUTHOR]

Storms can cause significant damage to forest areas, affecting biodiversity and infrastructure and leading to economic loss. Thus, rapid detection and mapping of windthrows are crucially important for forest management. Recent advances in computer vision have led to highly-accurate image classification algorithms such as Convolutional Neural Network (CNN) architectures. In this study, we tested and implemented an algorithm based on CNNs in an ArcGIS environment for automatic detection and mapping of damaged areas. The algorithm was trained and tested on a forest area in Bavaria, Germany.. It is a based on a modified U-Net architecture that was optimized for the pixelwise classification of multispectral aerial remote sensing data. The neural network was trained on labeled damaged areas from after-storm aerial orthophotos of a ca. 109 k m 2 forest area with RGB and NIR bands and 0.2-m spatial resolution. Around 10 7 pixels of labeled data were used in the process. Once the network is trained, predictions on further datasets can be computed within seconds, depending on the size of the input raster and the computational power used. The overall accuracy on our test dataset was 92 % . During visual validation, labeling errors were found in the reference data that somewhat biased the results because the algorithm in some instance performed better than the human labeling procedure, while missing areas affected by shadows. Our results are very good in terms of precision, and the methods introduced in this paper have several additional advantages compared to traditional methods: CNNs automatically detect high- and low-level features in the data, leading to high classification accuracies, while only one after-storm image is needed in comparison to two images for approaches based on change detection. Furthermore, flight parameters do not affect the results in the same way as for approaches that require DSMs and DTMs as the classification is only based on the image data themselves, and errors occurring in the computation of DSMs and DTMs do not affect the results with respect to the z component. The integration into the ArcGIS Platform allows a streamlined workflow for forest management, as the results can be accessed by mobile devices in the field to allow for high-accuracy ground-truthing and additional mapping that can be synchronized back into the database. Our results and the provided automatic workflow highlight the potential of deep learning on high-resolution imagery and GIS for fast and efficient post-disaster damage assessment as a first step of disaster management. [ABSTRACT FROM AUTHOR]

Crowd counting, which is widely used in disaster management, traffic monitoring, and other fields of urban security, is a challenging task that is attracting increasing interest from researchers. For better accuracy, most methods have attempted to handle the scale variation explicitly. which results in huge scale changes of the object size. However, earlier methods based on convolutional neural networks (CNN) have focused primarily on improving accuracy while ignoring the complexity of the model. This paper proposes a novel method based on a lightweight CNN-based network for estimating crowd counting and generating density maps under resource constraints. The network is composed of three components: a basic feature extractor (BFE), a stacked à trous convolution module (SACM), and a context fusion module (CFM). The BFE encodes basic feature information with reduced spatial resolution for further refining. Various pieces of contextual information are generated through a short pipeline in SACM. To generate a context fusion density map, CFM distills feature maps from the above components. The whole network is trained in an end-to-end fashion and uses a compression factor to restrict its size. Experiments on three highly-challenging datasets demonstrate that the proposed method delivers attractive performance. [ABSTRACT FROM AUTHOR]