Search

Image recognition has long been one of the research hotspots in computer vision tasks. The development of deep learning is rapid in recent years, and convolutional neural networks usually need to be designed with fixed resources. If sufficient resources are available, the model can be scaled up to achieve higher accuracy, for example, VggNet, ResNet, GoogLeNet, etc. Although the accuracy of large-scale models has been improved, the following problems will occur with the expansion of model scale: (1) There may be over-fitting; (2) increasing model parameters; (3) slow model convergence. This paper proposes a design method for a modular convolutional neural network model which solves the problem of over-fitting and large model parameters by connecting multiple modules in parallel. Moreover, each module contains several submodules (three submodules in this paper) and fuses the features extracted from the submodules. The model convergence can be accelerated by using the fused features (the fused features contain more image information). In this study, we add a gate unit based on the attention mechanism to the model, which aims to optimize the structure of the model (select the optimal number of modules), allowing the model to select an optimum network structure by learning and dynamically reducing FLOPs (floating-point operations per second) of the model. Compared to VggNet, ResNet, and GoogLeNet, the structure of the model proposed in this paper is simple and the parameters are small. The proposed model achieves good results in the Kaggle datasets Cats-vs.-Dogs (99.3%), 10-Monkey Species (99.26%), and Birds-400 (99.13%). [ABSTRACT FROM AUTHOR]

Deep learning networks have recently demonstrated yielded impressive progress for multi-exposure image fusion. However, how to restore realistic texture details while correcting color distortion is still a challenging problem to be solved. To alleviate the aforementioned issues, in this paper, we propose an attention-guided global-local adversarial learning network for fusing extreme exposure images in a coarse-to-fine manner. Firstly, the coarse fusion result is generated under the guidance of attention weight maps, which acquires the essential region of interest from both sides. Secondly, we formulate an edge loss function, along with a spatial feature transform layer, for refining the fusion process. So that it can take full use of the edge information to deal with blurry edges. Moreover, by incorporating global-local learning, our method can balance pixel intensity distribution and correct the color distortion on spatially varying source images from both image/patch perspectives. Such a global-local discriminator ensures all the local patches of the fused images align with realistic normal-exposure ones. Extensive experimental results on two publicly available datasets show that our method drastically outperforms state-of-the-art methods in visual inspection and objective analysis. Furthermore, sufficient ablation experiments prove that our method has significant advantages in generating high-quality fused results with appealing details, clear targets, and faithful color. Source code will be available at https://github.com/JinyuanLiu-CV/AGAL. [ABSTRACT FROM AUTHOR]

RGB and thermal infrared (RGBT) image saliency detection is a relatively new direction in the field of computer vision. Combining the advantages of RGB images and T images can significantly improve detection performance. Currently, there are only a few methods to work on RGBT saliency detection, and the number of image samples cannot meet the training requirements for deep learning, so it remains valuable to propose an effective unsupervised method. In this paper, we present an unsupervised RGBT saliency detection method based on multi-graph fusion and learning. Firstly, RGB images and T images are adaptively fused based on boundary information to produce more accurate superpixels. Next, a multi-graph fusion model is proposed to selectively learn useful information from multi-modal images. Finally, we implement the theory of finding good neighbors in the graph affinity and propose different algorithms for two stages of saliency ranking. Experimental results on three RGBT datasets show that the proposed method is effective compared with the state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

In recent years, visual object tracking, as a prominent research area in computer vision, has garnered significant attention. To bolster the robustness of trackers across a spectrum of complex scenarios, researchers actively explore the synergistic potential of visible and thermal infrared images, aiming to design more potent tracking systems. This paper presents a comprehensive review of target tracking technology based on visible and thermal infrared information, encompassing three key aspects. Firstly, we categorize existing RGBT tracking methods into two main categories: traditional-based methods and deep learning-based methods. This classification facilitates a systematic understanding and comparison of the strengths and weaknesses of different approaches, providing a solid foundation for future research. Secondly, we focus on the evolution of RGBT datasets and analyze the performance of diverse tracking methods on these datasets. Research in this domain aids in evaluating the applicability of existing methods in real-world scenarios and offers guidance for future dataset construction. Finally, we delve into future research directions from multiple perspectives, including model design and dataset construction. In terms of model design, researchers are encouraged to explore more efficient feature extraction methods and innovative model fusion structures to further enhance tracker performance. Regarding dataset construction, increased attention should be given to ensure diversity in real-world scenarios, guaranteeing optimal tracker performance across a variety of complex conditions. In conclusion, this review makes a comprehensive analysis of the development of RGBT tracking from different perspectives, and provides a valuable reference for researchers in related fields such as multi-modal tracking and image fusion. By systematically classifying and analyzing existing research while outlining future research prospects, this review aims to foster the continued development of this field and inspire the emergence of more innovative work. • Comprehensive review of RGBT tracking technology into two main categories. • Sufficient experimental results on several large RGBT benchmark datasets. • A prospective future discussion underlying model design and dataset construction. [ABSTRACT FROM AUTHOR]

• A comprehensive review of 3D Object Detection for driverless vehicles is provided. • A generic pipeline architecture is proposed to describe the models workflow. • Taxonomies are proposed to categorize the existing models regarding design choices. • A comparative study for models is conducted and the relevant resources released. • The current challenges and future trends of multi-focus image fusion are discussed. Autonomous vehicles are becoming central for the future of mobility, supported by advances in deep learning techniques. The performance of aself-driving system is highly dependent on the quality of the perception task. Developments in sensor technologies have led to an increased availability of 3D scanners such as LiDAR, allowing for a more accurate representation of the vehicle's surroundings, leading to safer systems. The rapid development and consequent rise of research studies around self-driving systems since early 2010, resulted in a tremendous increase in the number and novelty of object detection methods. After the first wave of works that essentially tried to expand known techniques from object detection in images, more recently there has been a notable development in newer and more adapted to LiDAR data works. This paper addresses the existing literature on object detection using LiDAR data within the scope of self-driving and brings a systematic way for analysing it. Unlike general object detection surveys, we will focus on point-cloud data, which presents specific challenges, notably its high-dimensional and sparse nature. This work introduces a common object detection pipeline and taxonomy to facilitate a thorough comparison between different techniques and, departing from it, this work will critically examine the representation of data (critical for complexity reduction), feature extraction and finally the object detection models. A comparison between performance results of the different models is included, alongside with some future research challenges. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) have facilitated substantial progress on various problems in computer vision and image processing. However, applying them to image fusion has remained challenging due to the lack of the labelled data for supervised learning. This paper introduces a deep image fusion network (DIF-Net), an unsupervised deep learning framework for image fusion. The DIF-Net parameterizes the entire processes of image fusion, comprising of feature extraction, feature fusion, and image reconstruction, using a CNN. The purpose of DIF-Net is to generate an output image which has an identical contrast to high-dimensional input images. To realize this, we propose an unsupervised loss function using the structure tensor representation of the multi-channel image contrasts. Different from traditional fusion methods that involve time-consuming optimization or iterative procedures to obtain the results, our loss function is minimized by a stochastic deep learning solver with large-scale examples. Consequently, the proposed method can produce fused images that preserve source image details through a single forward network trained without reference ground-truth labels. The proposed method has broad applicability to various image fusion problems, including multi-spectral, multi-focus, and multi-exposure image fusions. Quantitative and qualitative evaluations show that the proposed technique outperforms existing state-of-the-art approaches for various applications. [ABSTRACT FROM AUTHOR]

The existing classification methods for Panax notoginseng taproots suffer from low accuracy, low efficiency, and poor stability. In this study, a classification model based on image feature fusion is established for Panax notoginseng taproots. The images of Panax notoginseng taproots collected in the experiment are preprocessed by Gaussian filtering, binarization, and morphological methods. Then, a total of 40 features are extracted, including size and shape features, HSV and RGB color features, and texture features. Through BP neural network, extreme learning machine (ELM), and support vector machine (SVM) models, the importance of color, texture, and fusion features for the classification of the main roots of Panax notoginseng is verified. Among the three models, the SVM model performs the best, achieving an accuracy of 92.037% on the prediction set. Next, iterative retaining information variables (IRIVs), variable iterative space shrinkage approach (VISSA), and stepwise regression analysis (SRA) are used to reduce the dimension of all the features. Finally, a traditional machine learning SVM model based on feature selection and a deep learning model based on semantic segmentation are established. With the model size of only 125 kb and the training time of 3.4 s, the IRIV-SVM model achieves an accuracy of 95.370% on the test set, so IRIV-SVM is selected as the main root classification model for Panax notoginseng. After being optimized by the gray wolf optimizer, the IRIV-GWO-SVM model achieves the highest classification accuracy of 98.704% on the test set. The study results of this paper provide a basis for developing online classification methods of Panax notoginseng with different grades in actual production. [ABSTRACT FROM AUTHOR]