Search

In order to address the problem of decreased accuracy in vehicle object detection models when facing low-light conditions in nighttime environments, this paper proposes a method to enhance the accuracy and precision of object detection by using the image translation technology based on the Generative Adversarial Network (GAN) in the field of computer vision, specifically the CycleGAN, from the perspective of improving the training set of object detection models. This is achieved by transforming the existing well-established daytime vehicle dataset into a nighttime vehicle dataset. The proposed method adopts a comparative experimental approach to obtain translation models with different degrees of fitting by changing the training set capacity, and selects the optimal model based on the evaluation of the effect. The translated dataset is then used to train the YOLO-v5-based object detection model, and the quality of the nighttime dataset is evaluated through the evaluation of annotation confidence and effectiveness. The research results indicate that utilizing the translated nighttime vehicle dataset for training the object detection model can increase the area under the PR curve and the peak F1 score by 10.4% and 9% respectively. This approach improves the annotation accuracy and precision of vehicle object detection models in nighttime environments without requiring additional labeling of vehicles in monitoring videos. [ABSTRACT FROM AUTHOR]

Numerous people are interested in learning yoga due to the increased tension levels in the modern lifestyle, and there are a variety of techniques or resources available. Yoga is practiced in yoga centers, by personal instructors, and through books, the Internet, recorded videos, etc. As the aforementioned resources may not always be available, a large number of people will opt for self-study in fast-paced lifestyles. Self-learning makes it impossible to recognize an incorrect posture. Incorrect poses will have a negative effect on the patient’s health, causing severe agony and long-term chronic issues. Computer vision (CV)-related techniques derive pose features and conduct pose analysis using non-invasive CV methods. The application of machine learning (ML) and artificial intelligence (AI) techniques to an inter-disciplinary field like yoga becomes quite difficult. Due to its potent feature learning ability, deep learning (DL) has recently achieved an impressive level of performance in classifying yoga poses. In this paper, an artificial algae optimizer with hybrid deep learning-based yoga pose estimation (AAOHDL-YPE) model is presented. The presented AAOHDL-YPE model analyzes yoga video clips to estimate pose. Utilizing Part Confidence Map and Part Affinity Field with bipartite equivalent and parsing, OpenPose can be employed to determine the joint location. The deep belief network (DBN) model is then used for Yoga recognition. Finally, the AAO algorithm is utilized to enhance the EfficientNet model’s recognition performance. The results of a comprehensive experimentation analysis reveal that the AAOHDL-YPE technique produces superior results in comparison to existing methods. [ABSTRACT FROM AUTHOR]

In this paper, 1,538 papers retrieved with the keywords "sports artificial intelligence (AI)" on the Web of Science database since 2007 were taken as the data source, and the Cite Space V software was used to visualize and analyze them. A visual knowledge graph was used to streamline the countries, institutions and authors conducting sports AI research, discipline distribution, research hotspots and development trends in the past 15 years. Subsequently, its development direction and research progress were discussed. Sports AI was widely distributed, with the US, China and the UK leading the way. The most prolific authors and teams in research on sports AI were concentrated in American universities. Their main research direction is to develop and improve smart wearable devices based on machine learning and deep learning technologies for different groups of people. Research on sports AI involved multiple disciplines, which mainly applied and referred to research methodologies and theories on engineering, computer science and sports science. It could be seen from the frequency and centrality of keywords that in the current field of sports AI, machine learning is the main direction, artificial neural networks is the main algorithm, and practical and empirical research based on data mining is the focus. The research hotspots were divided into three major clusters: physical health promotion, sports injury prevention and control, and athletic performance enhancement. How to introduce intelligent technology into sports for a perfect integration still has an arduous and long way to go. Future development requires joint efforts and participation of scientific researchers, professionals and common people. [ABSTRACT FROM AUTHOR]

The need for a monitoring system has grown as a result of rising crime and anomalous activity. To avoid unusual incidents, the common man initiated video surveillance of important areas, which was then passed on to the government. In typical surveillance operations, surveillance devices create a vast volume of data that must be manually analysed. Manually handling huge data sets in real time results in information loss. To prevent abnormal incidents, the actions in sensitive areas can be properly monitored, evaluated, and alerted to the appropriate authorities. Previous deep learning-based activity identification methods have appeared, but the findings are inaccurate, and the proposed Hybrid Machine Learning Algorithms (HMLA) incorporate two detection methods for surveillance videos like as Transfer Learning (TL) and Continual Learning (CL). As a result, the suspicious activity in the video may be missed. Consequently, numerous image processing and computer vision technologies were used in activity detection to decrease human effort and mistakes in surveillance operations. Activities in sensitive areas can be properly monitored and evaluated to avoid unusual incidents, and the appropriate authorities may be alerted. Hence, in order to decrease human error and effort in surveillance operations, activity recognition embraced a variety of image processing and computer vision technologies. In this present work, the capacity has constraints that impact recognition accuracy. Consequently, this research paper presents a HMLA based technique that uses feature extraction using multilayer (Long Short Term Memory) LSTM, Convolutional Neural Networks (CNN), and Temporal feature extraction using multilayer LSTM to improve identification accuracy by 96% while requiring minimal execution time. To show the superior performance of the proposed hybrid machine learning technique, a standard UCF crime dataset was utilised for experimental analysis and compared to existing deep learning algorithms. [ABSTRACT FROM AUTHOR]

Computer vision algorithms based on deep learning have evolved to play a major role in sports analytics. Nevertheless, in sports like table tennis, detecting the ball is a challenge as the ball travels at a high velocity. However, the events in table tennis games can be detected and classified by obtaining the locations of the ball. Therefore, existing methodologies predict the trajectories of the ball but do not detect and classify the in-game events. This paper, therefore, proposes a ball detection and trajectory analysis (BDTA) approach to detect the location of the ball and predict the trajectory to classify events in a table tennis game. The proposed methodology is composed of two parts: i) Scaled-YOLOv4 which can detect the precise position of the ball ii) Analysis of trajectory based on ball coordinates to detect and classify the events. The dataset was prepared and labeled as a ball after enhancing the frame resolution with a super-resolution technique to get the accurate position of the ball. The proposed approach demonstrates 97.8% precision and 98.1% f1-score in detecting the location of the ball and 97.47% precision and achieved 97.8% f-score in classifying in-game events. [ABSTRACT FROM AUTHOR]

This paper proposes that the task of single-image low-light enhancement can be accomplished by a straightforward method named Opt2Ada. It contains a series of pixel-level operations, including an optimized illuminance channel decomposition, an adaptive illumination enhancement, and an adaptive global scaling. Opt2Ada is traditional and it does not rely on architecture engineering, super-parameter tuning, or specific training dataset. Its parameters are generic and it has better generalization capability than existing data-driven methods. For evaluation, both the full-reference, non-reference, and semantic metrics are calculated. Extensive experiments on real-world low-light images demonstrate the superiority of Opt2Ada over recent traditional and deep learning algorithms. Due to its flexibility and effectiveness, Opt2Ada can be deployed as a pre-processing subroutine for high-level computer vision applications. [ABSTRACT FROM AUTHOR]

Human-wildlife conflicts in the habitats along the forest fringes are a substantial issue. An automated monitoring system that can find animal breaches and deter them from foraging fields is essential to solve this conflict. However, automatically forefending the intruding animals is a challenging task. In this paper, we propose a deep learning model for elephant identification using YOLO lite with knowledge distillation which could be easily deployed in edge devices. We also propose an elephant re-identification system using Siamese network which is helpful in tracking the number of times the elephant tries to forage the field. This re-encounter information about the same elephant can be used to decide the averting sound for the particular elephant. The proposed system is found to show an accuracy of 89%, which is provides good performance improvement when compared to the state of art models proposed for animal identification. Thus the proposed lite weight knowledge distillation based animal identification model and deep learning based animal re-identification model can be employed in edge devices for real time monitoring and animal deterring to safe guard the farm fields. [ABSTRACT FROM AUTHOR]

As an advanced machine vision task, traffic sign recognition is of great significance to the safe driving of autonomous vehicles. Haze has seriously affected the performance of traffic sign recognition. This paper proposes a dehazing network, including multi-scale residual blocks, which significantly affects the recognition of traffic signs in hazy weather. First, we introduce the idea of residual learning, design the end-to-end multi-scale feature information fusion method. Secondly, the study used subjective visual effects and objective evaluation metrics such as Visibility Index (VI) and Realness Index (RI) based on the characteristics of the real-world environment to compare various traditional dehazing and deep learning dehazing method with good performance. Finally, this paper combines image dehazing and traffic sign recognition, using the algorithm of this paper to dehaze the traffic sign images under real-world hazy weather. The experiments show that the algorithm in this paper can improve the performance of traffic sign recognition in hazy weather and fulfil the requirements of real-time image processing. It also proves the effectiveness of the reformulated atmospheric scattering model for the dehazing of traffic sign images. [ABSTRACT FROM AUTHOR]

Rain streaks severely affect the perception of the content and structure of an image so high-performance deraining algorithms are needed in order to eliminate the effects of various rain streaks for high-level computer vision tasks. Although much progress has been made with existing deraining methods, the task of single image deraining remains challenging. In this paper, we first point out that existing Transformers lack sufficient ability to capture channel attention which restricted the ability of models in deraining. To improve the performance of deraining model, we propose a dual branch deraining network based on Transformer. One branch uses dense connections to connect Transformer modules which embed the attention of a composite channel. This branch captures channel attention more finely to learn the representation of rain streaks features. The other branch first obtains features at different scales by gradually expanding the receptive field, then uses these features to obtain attention for regional features, and finally uses the attention to guide the model to focus on areas of high rain streaks density and large scales. By fusing these two branches, the model is able to capture channel attention more finely and to focus on regions of high rain streaks density and large scales. The extensive experimental results on synthetic and real datasets demonstrate that the proposed method outperforms most advanced deraining methods. [ABSTRACT FROM AUTHOR]

Human Activity Recognition (HAR) is the most popular research area in the pervasive computing field in recent years. Sensor data plays a vital role in identifying several human actions. Convolutional Neural Networks (CNNs) have now become the most recent technique in the computer vision phenomenon, but still, it is premature to use CNN for sensor data, particularly in ubiquitous and wearable computing.Deep CNN requires huge dataset and models which increases the computational complexity. Transfer learning that uses the pre trained CNNwith fine tuning is the better alternative to reduce the training cost.In this paper, we have proposed the idea of transforming the raw accelerometer and gyroscope sensor data to the visual domain by using our novel activity image creation method (NAICM). Pre-trained CNN (AlexNet) has been used on the converted image domain information. The proposed method is evaluated on several online available human activity recognition dataset. The results show that the proposed novel activity image creation method (NAICM) has successfully created the activity images with a classification accuracy of 98.36% using pre trained CNN. [ABSTRACT FROM AUTHOR]

A computational method for detecting pineapple ripening could lead to increased agricultural productivity. It is possible to predict fruit maturity before harvesting to increase agricultural productivity. A ripe fruit's quality, its standard content of physical and chemical properties will increase the value of a good when traded outside the market. This paper studies and improves the Tiny YOLO-v4 model for identifying the pineapple ripening period. Researchers studied pineapples in a pineapple garden in Vietnam's central region. They wanted to determine when pineapples were ripe. The API and the website are based on the YOLO innovation model. Apps and website APIs will be available for mobile devices so that people can monitor fruits. Technology transfer and academic research are combined in this study. We prepared the pineapple data set by using 5,000,000 pineapples harvested from the pineapple farm at different stages of growth. To make the measurements, we improved the YOLO-v4 algorithm. This results in a more accurate training model and reduced train-ing time. A 98.26% recognition accuracy is quite impressive. Research takes place at large-scale plantations, so the models are created from the data collected at the plantations and are used as labels; training takes a long time for the tiniest details about pineapples, and finding pineapple-growing regions takes a long time. The deep learning classifier was able to process pineapple plantation photos by using the camera on the mobile phone. [ABSTRACT FROM AUTHOR]

Universal Adversarial Perturbations(UAPs), which are image-agnostic adversarial perturbations, have been demonstrated to successfully deceive computer vision models. Proposed UAPs in the case of data-dependent, use the internal layers' activation or the output layer's decision values as supervision. In this paper, we use both of them to drive the supervised learning of UAP, termed as fully supervised UAP(FS-UAP), and design a progressive optimization strategy to solve the FS-UAP. Specifically, we define an internal layers supervised objective relying on multiple major internal layers' activation to estimate the deviations of adversarial examples from legitimate examples. We also define an output layer supervised objective relying on the logits of output layer to evaluate attacking degrees. In addition, we use the UAP found by previous stage as the initial solution of the next stage so as to progressively optimize the UAP stage-wise. We use seven networks and ImageNet dataset to evaluate the proposed FS-UAP, and provide an in-depth analysis for the latent factors affecting the performance of universal attacks. The experimental results show that our FS-UAP (i) has powerful capability of cheating CNNs (ii) has superior transfer-ability across models and weak data-dependent (iii) is appropriate for both untarget and target attacks. [ABSTRACT FROM AUTHOR]

In this paper, the authors propose an improved convolutional neural network for automatic arrhythmia classification using Electro-Cardio-Gram (ECG) signal. It is essential to periodically monitor the heart beat arrhythmia to reduce the risk of death due to cardiovascular disease (CVD). The Visual Geometry Group network (VGGNet) is being widely used in computer vision problems. However, the same network cannot be used for classification of ECG beats as ECG signal is different from image signal in terms of dimensionality and inherent features. Thus, the authors investigated the effect of decreasing depth and width of a convolutional neural network in context of cardiac arrhythmia classification. In this paper, six configurations differing in depth and width are evaluated using benchmark MIT-BIH database. A deep network having thirteen convolution layers but with a smaller number of filters (reduced width) showed outstanding performance for the given problem. Based on the findings, the work is further extended to propose an improved convolutional neural network named as modified VGGNet (mVGGNet) for the task of ECG arrhythmia classification into four classes which are normal (N), ventricular ectopic beat (VEB), supraventricular ectopic beat (SVEB) and fusion beat (F). The hyper-parameters of the proposed architecture were optimized using sequential model based global optimization (SMBO) algorithm. The proposed architecture is evaluated using subjected-oriented patient-independent evaluation protocol. The performance is evaluated using five-fold cross-validation. The proposed mVGGNet achieved 98.79% and 99.16% accuracy for ventricular ectopic beats (VEB) and supraventricular ectopic beats (SVEB) classification respectively. The proposed method resulted in higher specificity and precision as compared to other state-of-the-art algorithms. Thus, it can be effectively used for ECG arrhythmia classification. [ABSTRACT FROM AUTHOR]

As an important branch of Nature Language Processing (NLP), how to extract useful text information and effective long-range associations has always been a bottleneck for text classification. With the great effort of deep learning researchers, deep Convolutional Neural Networks (CNNs) have made remarkable achievements in Computer Vision but still controversial in NLP tasks. In this paper, we propose a novel deep CNN named Deep Pyramid Temporal Convolutional Network (DPTCN) for short text classification, which is mainly consisting of concatenated embedding layer, causal convolution, 1/2 max pooling down-sampling and residual blocks. It is worth mentioning that our work was highly inspired by two well-designed models: one is temporal convolutional network for sequential modeling; another is deep pyramid CNN for text categorization; as their applicability and pertinence remind us how to build a model in a special domain. In the experiments, we evaluate the proposed model on 7 datasets with 6 models and analyze the impact of three different embedding methods. The results prove that our work is a good attempt to apply word-level deep convolutional network in short text classification. [ABSTRACT FROM AUTHOR]

In the last decade, there have been extensive reports of world health organization (WHO) on breast cancer. About 2.1 million women are affected every year and it is the second most leading cause of cancer death in women. Initial detection and diagnosis of cancer appreciably increase the chance of saving lives and reduce treatment costs. In this paper, we perform a survey of the techniques utilized in breast cancer detection and diagnosis in image processing, machine learning (ML), and deep learning (DL). We also proposed a novel computer-vision based cost-effective method for breast cancer detection and diagnosis. Along with the detection and diagnosis of breast cancer, our proposed method is capable of finding the exact position of the abnormality present in the breast that will help in breast-conserving surgery or partial mastectomy. The proposed method is the simplest and cost-effective approach that has produced highly accurate and useful outcomes when compared with the existing approach. [ABSTRACT FROM AUTHOR]

The government has established special schools to cater to the needs of children with disabilities but they are often segregated rather than receiving equitable opportunities. Artificial Intelligence has opened new ways to promote special education with advanced learning tools. These tools enable to adapt to a typical classroom set up for all the students with or without disabilities. To ensure social equity and the same classroom experience, a coherent solution is envisioned for inclusive education. This paper aims to propose a cost-effective and integrated Smart Learning Assistance (SLA) tool for Inclusive Education using Deep Learning and Computer Vision techniques. It comprises speech to text and sign language conversion for hearing impaired students, sign language to text conversion for speech impaired students, and Braille to text for communicating with visually impaired students. The tool assists differently-abled students to make use of various teaching-learning opportunities conferred to them and ensures convenient two-way communication with the instructor and peers in the classroom thus makes learning easier. [ABSTRACT FROM AUTHOR]

Data cognition plays an important role in cognitive computing. Cognition of low-resolution (LR) image is a long-stand problem because LR images have insufficient information about objects. For better cognition of LR images, a multi-resolution residual network (MRRN) is proposed to improve image resolution in this paper for cognitive computing systems. In MRRN, a multi-resolution feature learning (MRFL) strategy is introduced to achieve satisfying performance with low computational costs. Inspired by image pyramids, a feature pyramid is designed to implement multi-resolution feature learning in the building unit of the proposed MRRN. Specifically, multi-resolution residual units (MRRUs) are introduced as the building units of the proposed network, which consist of a feature pyramid decomposition stage and a feature reconstruction stage. To obtain informative features, transferred skip links (TSLs) are utilized to transfer fine-grain residual features in the pyramid decomposition stage to the reconstruction stage. The effectiveness of MRFL and TSL is demonstrated by ablation experiments. Also, the tests on standard benchmarks indicate the superiority of the proposed MRRN over other state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Deep learning is a field of Artificial Intelligence that has recently drawn a lot of attention with the desire to build up a quick, automatic and accurate system for image identification and classification. Deep learning serves as a fundamental part of modern computer vision solutions. However, as the architectures become deep and powerful new challenges in the process of training emerge. This includes the computational cost associated with training deep and large networks. In this work, the focus is on pruning and evaluation of state-of-the-art deep convolutional neural network for image-based plant disease and plants species classification. Pruning filters allow the reduction of parameters by removing unimportant filters and its feature maps. In this paper, the performance of pruned networks is evaluated across three datasets. It is observed that pruned DenseNet with Self-Normalization Neural Network (SNN) approach learns 2x faster compared to the initial DenseNet architecture. Additionally, pruning filters allow the reduction of the number of parameters and FLOPs by approximately 14% and 25% respectively. The aim is to create a fast and efficient model for the purpose of identification of plant diseases. Fast methods are desired for early identifications of diseases before damages occur. The proposed method achieves a satisfactory accuracy performance on PlantVillage, LeafSnap and Swedish-leaf dataset using held-out dataset. Our best pruned model gives an accuracy of 99.24%, 86.64%, and 97.5% on PlantVillage, LeafSnap, and Swedish-leaf datasets respectively. [ABSTRACT FROM AUTHOR]

In order to promote the development of visual target tracking technology, this paper uses discrete differential algorithm in computer to assist analysis. First, the development status of target tracking technology is expounded, constructing the target tracking model by using different image sequences. In order to enhance the efficiency of target tracking, discrete differential algorithm is added. The algorithm can reduce the complexity of the computation and adopt the traditional optical flow evaluation method and level set frame tracking method to perfect. An optimization and update scheme is proposed from the algorithm flow and evaluation model, and an effective evaluation model is established by effectively evaluating the state of the visual tracking target model. In the test of target tracking model, through the comparison of different targets and the efficiency of data calculation, it is proved that the model can be layered for the target, and the addition of discrete differential algorithm effectively solves the problem of scale self-adaptation, which deserves further popularization. [ABSTRACT FROM AUTHOR]

Deep neural networks have dramatically gained immense potential in almost every field like computer vision, natural language processing, biomedical informatics etc. Among these networks, autoencoders are popular in performing dimensionality reduction task, while learning a representation for an unlabeled dataset. A usual way of dealing with such networks is to pre-train them in a layer-wise fashion, and consequently fine-tune the whole stack in a supervised manner. In this paper, a pair-wise training strategy is proposed to determine optimum model parameters by reducing training time as well as the complexity of training a convolutional autoencoder without compromising its accuracy. The proposed approach works in a fully unsupervised manner and has been tested on datasets like MNIST, CIFAR10, and CIFAR100 and it shows that the training time has improved by an average of 25% on these three datasets. [ABSTRACT FROM AUTHOR]

Visual tracking is a very challenging task in computer vision. In this paper, we present a general-purpose framework for robust tracking. We propose to couple one-shot learning and online discriminative learning together to address the fundamental stability-plasticity issue for tracking. A one-shot learner through offline training on large-scale datasets is used as a stable detector which does not suffer model drift while an online discriminative learner is adopted as the tracker which is adaptive to significant appearance changes. Based on the directive framework, we design a baseline tracking model to verify its effectiveness. In practice, a deep Siamese network trained offline plays as the one-shot learner which can re-detect the target in case of tracking drift and failure. A correlation classifier which incorporates a translation model and a scale model plays as the online learner. Through the coupling of the offline and online learning, the simple baseline tracker achieves a good balance between stability and adaptivity without time-consuming optimization. Experimental results on the large-scale benchmark dataset demonstrate the effectiveness of the proposed framework within which the designed baseline tracker outperforms many state-of-the-art methods both in precision and robustness. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) are important methods in deep learning. They have presented up-to-date performance in different challenging areas, such as natural language processing and computer vision. These powerful and efficient neural networks implement training slowly under a massive number of network training parameters. The primary challenge is to reduce the training time for large volumetric data. CNNs have a small number of parameters at convolutional layers and a large number of parameters at fully connected layers. Training time can be reduced by the use of computing parallelism according to the characteristics of CNN layers. This paper presents an optimized parallelism algorithm using a communication strategy for CNN training in distributed graphic processing units (GPUs). We use the butterfly reduction communication strategy and apply data and model parallelisms at convolutional and fully connected layers respectively. A model is divided among distributed GPUs, and each division of the model works according to the characteristics of CNNs. This hybrid parallel approach is more desirable than previous parallelism alternatives, such as data parallelism only and model parallelism alone, which have been applied to modern CNNs. Experimental results reveal that this parallel approach with butterfly communication strategy can enhance accuracy and decrease training time. [ABSTRACT FROM AUTHOR]