Search

In action recognition that relies on visual information, activities are recognized through spatio-temporal features from different modalities. The challenge of temporal modeling has been a long-standing issue in this field. There are a limited number of methods, such as pre-computed motion features, three-dimensional (3D) filters, and recurrent neural networks (RNNs), that are used in deep-based approaches to model motion information. However, the success of transformers in modeling long-range dependencies in natural language processing tasks has recently caught the attention of other domains, including speech, image, and video, as they can rely entirely on self-attention without using sequence-aligned RNNs or convolutions. Although the application of transformers to action recognition is relatively new, the amount of research proposed on this topic in the last few years is impressive. This paper aims to review recent progress in deep learning methods for modeling temporal variations in multimodal human action recognition. Specifically, it focuses on methods that use transformers for temporal modeling, highlighting their key features and the modalities they employ, while also identifying opportunities and challenges for future research. [ABSTRACT FROM AUTHOR]

Scene classification is one of the active research domains of artificial intelligence (AI) with many real-world applications. This paper presents a new scene classification approach based on the Belief Function Theory, which provides a more effective way of handling uncertainty information compared to traditional probability-based methods. Unlike previous methods that rely on probabilities, which have proved their limitations, the main contribution of our approach is the use of belief degrees to classify unknown scenes based on object labels. We conduct experiments on three well-known datasets (SUN397, MIT Indoor, and LabelMe) and compare our results with state-of-the-art methods. Our approach achieves competitive results with a simple and robust framework that outperforms previous methods in some cases. We also provide insights into the strengths and limitations of our approach and discuss potential future directions for research. Overall, our work demonstrates the effectiveness of the Belief Function theory in scene classification and opens up new avenues for further research and innovation in this area. [ABSTRACT FROM AUTHOR]

Deep learning-based object detection has become popular due to its strong learning ability and advantages in dealing with occlusion, scale transformation, and context changes. In recent years, it has become a research hotspot. This paper presents the current Deep Learning models from Generic and Salient detection models ranging from one-stage to two-stage for multi-object detection in various applications. Nevertheless, we also examined the advantages and some drawbacks of those models. Furthermore, challenges such as variation in object scales, computation time, illumination differing from various applications, and promising research directions of Deep Learning models are discussed. Finally, we proposed Dense PRediction Simplified (DPRS) based on the YOLO model. Backbones play a vital role in enhancing the performance of detection models, and efficient Backbone architecture will be fused to achieve the competitive state-of-art result. [ABSTRACT FROM AUTHOR]

Automatic detection, localization and interpretation of an unusual event in a sequence of video is a challenging task due to its equivocal and complex nature. The development of deep neural networks have paved the way for more efficient recognition and analysis of anomalous events in video data. With the introduction of convolutional neural network (CNN) and Long short-term memory (LSTM), the spatial and temporal features extraction became easier. In this paper, we propose an end-to-end trainable Inter-fused Autoencoder (IFA) which is designed using the assemblage of CNN and LSTM layers to detect the unwonted events in a video sequence. The proposed architecture is capable of exploiting both the spatial and temporal variation of video data. The reconstruction error is computed in terms of both MSE and PSNR for each testing video. A comparison is also carried out between MSE and PSNR to show that which assessment technique is better for a reconstructive model for recreating the video sequence. A well-optimized threshold is calculated which decides the fate of testing event i.e. either usual or unusual event. Using benchmark datasets, multiple experiments were carried out to demonstrate the efficacy of proposed architecture. [ABSTRACT FROM AUTHOR]

In previous years, the world has gone through several natural disasters like Tsunami, earthquakes, floods, tornadoes, hurricanes, cyclones, etc., and manufactured disasters such as stampedes, fire, terror attacks, etc. A large number of causalities are reported, with a massive loss to life, economy, and other things. Knowing this, we should make a transit from a reliable and flexible disaster management approach to a proactive one by leveraging advances in science and technology. A colossal increase in worldwide population points out that the occurrence of a crowd at any place is becoming more and more familiar with each passing day. It is undeniable that these mass gatherings often become a source of a crowd-related catastrophe such as sudden escape, terror attacks, mob lynching, human stampede, or human crushing. Prior research on the crowd's social, psychological, and computational dynamics has indicated that the crowd's behavior under such devastating conditions is greatly decisive for crowd safety, its access or escape from the affected region, and emergency evacuation. Despite this, there is a certain paucity of pragmatic research on extreme crowd-related use cases and how to deal with such situations effectively. Through the past years, people and media have shared the details of such happenings and their experiences on a micro-level through various social network mediums. Attempts are being made to analyze this data using advanced technological tools and methods to extract the trends out of such happenings and predict any future happenings so that countermeasures can be taken and they can be prevented. This paper makes a structured literature assessment on the current scenario and systematically surveys the studies made in this field. It paves the path for future rendezvous in this area to unearth the hidden gold mine of information along the timeline. Also, an attempt is made to develop a technological solution or system that may help achieve an elevated level of social security via holistic video surveillance capable of detecting any crowd-related anomaly and proactively warning the concerned authorities about any such casualty. This will ensure that crowd disasters can be prevented well in time by gaining prior insights about them. The system is developed that encompasses everything from human detection, tracking, and counting to any abnormal behavior detection. The same has been achieved with 93.33% accuracy. [ABSTRACT FROM AUTHOR]

Due to the increase in surveillance systems, there is a massive increase in surveillance data. As of now, the key challenge for video surveillance systems is analyzing these large video clips. It, therefore, has an enormous demand for intelligent video analysis systems capable of identifying activities and events. Since many researchers have emphasized the role of contextual knowledge and how the performance of video content analysis has improved in several ways, we have looked at different approaches in this study that can extract semantic information to human-level perception in the video. We also addressed open problems in semantics that come from event detection and irregular activity detection. Most methods/models are too coarse to accurately extract a complete set of information. Thus, we need to use a machine-readable format to view, process, store and extract meaningful information from the video data. In this paper, we discussed the methods/approaches for extracting low-level features, mid-level features, and high-level video features and their representation using Semantic Technologies. A taxonomy of hierarchical feature generation approaches is also provided. Some evaluation metrics for evaluating video activity and measuring the performance of the extraction features are explored. Community-approved benchmark datasets are also thoroughly surveyed and presented. The paper provides a complete framework of video research to develop an intelligent surveillance system. [ABSTRACT FROM AUTHOR]

Real-time detection of humans is an evolutionary research topic. It is an essential and prominent component of various vision based applications. Detection of humans in real-time video sequences is an arduous and challenging task due to various constraints like cluttered environment, occlusion, noise, etc. Many researchers are doing their research in this area and have published the number of researches so far. Determining humans in visual monitoring system is prominent for different types of applications like person detection and identification, fall detection for an elder person, abnormal surveillance, gender classification, crowd analysis, person gait characterization, etc. The main objective of this paper is to provide a comprehensive survey of the various challenges and modern developments seen for human detection methodologies in day vision. This paper consists of an overview of different human detection techniques and their classification based on various underlying factors. The algorithmic technicalities with their applicability to these techniques are deliberated in detail in the manuscript. Different humanitarian imperative factors have also been highlighted for comparative analysis of each human detection methodology. Our survey shows the difference between current research and future requirements. [ABSTRACT FROM AUTHOR]

Advanced intelligent surveillance systems are able to automatically analyze video of surveillance data without human intervention. These systems allow high accuracy of human activity recognition and then a high-level activity evaluation. To provide such features, an intelligent surveillance system requires a background subtraction scheme for human segmentation that captures a sequence of images containing moving humans from the reference background image. This paper proposes an alternative approach for human segmentation in videos through the use of a deep convolutional neural network. Two specific datasets were created to train our network, using the shapes of 35 different moving actors arranged on background images related to the area where the camera is located, allowing the network to take advantage of the entire site chosen for video surveillance. To assess the proposed approach, we compare our results with an Adobe Photoshop tool called Select Subject, the conditional generative adversarial network Pix2Pix, and the fully-convolutional model for real-time instance segmentation Yolact. The results show that the main benefit of our method is the possibility to automatically recognize and segment people in videos without constraints on camera and people movements in the scene (Video, code and datasets are available at http://graphics.unibas.it/www/HumanSegmentation/index.md.html). [ABSTRACT FROM AUTHOR]

Human activity recognition (HAR) systems attempt to automatically identify and analyze human activities using acquired information from various types of sensors. Although several extensive review papers have already been published in the general HAR topics, the growing technologies in the field as well as the multi-disciplinary nature of HAR prompt the need for constant updates in the field. In this respect, this paper attempts to review and summarize the progress of HAR systems from the computer vision perspective. Indeed, most computer vision applications such as human computer interaction, virtual reality, security, video surveillance and home monitoring are highly correlated to HAR tasks. This establishes new trend and milestone in the development cycle of HAR systems. Therefore, the current survey aims to provide the reader with an up to date analysis of vision-based HAR related literature and recent progress in the field. At the same time, it will highlight the main challenges and future directions. [ABSTRACT FROM AUTHOR]