Scene understanding plays a crucial role in autonomous driving by utilizing sensory data for contextual information extraction and decision making. Beyond modeling advances, the enabler for vehicles to become aware of their surroundings is the availability of visual sensory data, which expand the vehicular perception and realizes vehicular contextual awareness in real-world environments. Research directions for scene understanding pursued by related studies include person/vehicle detection and segmentation, their transition analysis, lane change, and turns detection, among many others Unfortunately, these tasks seem insufficient to completely develop fully-autonomous vehicles i.e. achieving level-5 autonomy, travelling just like human-controlled cars. This latter statement is among the conclusions drawn from this review paper: scene understanding for autonomous driving cars using vision sensors still requires significant improvements. With this motivation, this survey defines, analyzes, and reviews the current achievements of the scene understanding research area that mostly rely on computationally complex deep learning models. Furthermore, it covers the generic scene understanding pipeline, investigates the performance reported by the state-of-the-art, informs about the time complexity analysis of avant garde modeling choices, and highlights major triumphs and noted limitations encountered by current research efforts. The survey also includes a comprehensive discussion on the available datasets, and the challenges that, even if lately confronted by researchers, still remain open to date. Finally, our work outlines future research directions to welcome researchers and practitioners to this exciting domain.

Due to the rapid development of artificial intelligence technology, industrial sectors are revolutionizing in automation, reliability, and robustness, thereby significantly increasing quality and productivity. Most of the surveillance and industrial sectors are monitored by visual sensor networks capturing different surrounding environment images. However, during tempestuous weather conditions, the visual quality of the images is reduced due to contaminated suspended atmospheric particles that affect the overall surveillance systems. To tackle these challenges, this article presents a computationally efficient lightweight convolutional neural network referred to as Light-DehazeNet (LD-Net) for the reconstruction of hazy images. Unlike other learning-based approaches, which separately measure the transmission map and the atmospheric light, our proposed LD-Net jointly estimates both the transmission map and the atmospheric light using a transformed atmospheric scattering model. Furthermore, a color visibility restoration method is proposed to evade the color distortion in the dehaze image. Finally, we conduct extensive experiments using synthetic and natural hazy images. The quantitative and qualitative evaluation on different benchmark hazy datasets verify the superiority of the proposed method over other state-of-the-art image dehazing techniques. Moreover, additional experimentation validates the applicability of the proposed method in the object detection tasks. Considering the lightweight architecture with minimal computational cost, the proposed system is encouraged to be incorporated as an integral part of the vision-based monitoring systems to improve the overall performance.

Human activity recognition is an emerging and important area in computer vision which seeks to determine the activity an individual or group of individuals are performing. The applications of this field ranges from generating highlight videos in sports, to intelligent surveillance and gesture recognition. Most activity recognition systems rely on a combination of convolutional neural networks (CNNs) to perform feature extraction from the data and recurrent neural networks (RNNs) to determine the time dependent nature of the data. This paper proposes and designs two transformer neural networks for human activity recognition: a recurrent transformer (ReT), a specialized neural network used to make predictions on sequences of data, as well as a vision transformer (ViT), a transformer optimized for extracting salient features from images, to improve speed and scalability of activity recognition. We have provided an extensive comparison of the proposed transformer neural networks with the contemporary CNN and RNN-based human activity recognition models in terms of speed and accuracy for four publicly available human action datasets. Experimental results reveal that the proposed ViT-ReT framework attains a speedup of 2× over the baseline ResNet50-LSTM approach while attaining nearly the same level of accuracy. Furthermore, results show that the proposed ViT-ReT framework attains significant improvements over the state-of-the-art human action recognition methods in terms of both model accuracy and runtime for each of the datasets used in our experiments, thus verifying the suitability of the proposed ViT-ReT framework for human activity recognition in resource-constrained and real-time environments.

Vision-based human activity recognition (HAR) has emerged as one of the essential research areas in video analytics. Over the last decade, numerous advanced deep learning algorithms have been introduced to recognize complex human actions from video streams. These deep learning algorithms have shown impressive performance for the video analytics task. However, these newly introduced methods either exclusively focus on model performance or the effectiveness of these models in terms of computational efficiency, resulting in a biased trade-off between robustness and computational efficiency in their proposed methods to deal with challenging HAR problem. To enhance both the accuracy and computational efficiency, this paper presents a computationally efficient yet generic spatial–temporal cascaded framework that exploits the deep discriminative spatial and temporal features for HAR. For efficient representation of human actions, we propose an efficient dual attentional convolutional neural network (DA-CNN) architecture that leverages a unified channel–spatial attention mechanism to extract human-centric salient features in video frames. The dual channel–spatial attention layers together with the convolutional layers learn to be more selective in the spatial receptive fields having objects within the feature maps. The extracted discriminative salient features are then forwarded to a stacked bi-directional gated recurrent unit (Bi-GRU) for long-term temporal modeling and recognition of human actions using both forward and backward pass gradient learning. Extensive experiments are conducted on three publicly available human action datasets, where the obtained results verify the effectiveness of our proposed framework (DA-CNN+Bi-GRU) over the state-of-the-art methods in terms of model accuracy and inference runtime across each dataset. Experimental results show that the DA-CNN+Bi-GRU framework attains an improvement in execution time up to 167× in terms of frames per second as compared to most of the contemporary action-recognition methods.

Image classification is an important application for deep learning. With the advent of quantum technology, quantum neural networks (QNNs) have become the focus of research. Traditional deep learning-based image classification involves using a convolutional neural network (CNN) to extract features from the image and a multi-layer perceptron (MLP) network to create the decision boundaries. However, quantum circuits with parameters can extract rich features from images and also create complex decision boundaries. This paper proposes a hybrid QNN (H-QNN) model designed for binary image classification that capitalizes on the strengths of quantum computing and classical neural networks. Our H-QNN model uses a compact, two-qubit quantum circuit integrated with a classical convolutional architecture, making it highly efficient for computation on noisy intermediate-scale quantum (NISQ) devices that are currently leading the way in practical quantum computing applications. Our H-QNN model significantly enhances classification accuracy, achieving a 90.1% accuracy rate on binary image datasets. In addition, we have extensively evaluated baseline CNN and our proposed H-QNN models for image retrieval tasks. The obtained quantitative results exhibit the generalization of our H-QNN for downstream image retrieval tasks. Furthermore, our model addresses the issue of overfitting for small datasets, making it a valuable tool for practical applications.

In recent years, 360° videos have gained the attention of researchers due to their versatility and applications in real-world problems. Also, easy access to different visual sensor kits and easily deployable image acquisition devices have played a vital role in the growth of interest in this area by the research community. Recently, several 360° panorama generation systems have demonstrated reasonable quality generated panoramas. However, these systems are equipped with expensive image sensor networks where multiple cameras are mounted in a circular rig with specific overlapping gaps. In this paper, we propose an economical 360° panorama generation system that generates both mono and stereo panoramas. For mono panorama generation, we present a drone-mounted image acquisition sensor kit that consists of six cameras placed in a circular fashion with optimal overlapping gap. The hardware of our proposed image acquisition system is configured in such way that no user input is required to stitch multiple images. For stereo panorama generation, we propose a lightweight, cost-effective visual sensor kit that uses only three cameras to cover 360° of the surroundings. We also developed stitching software that generates both mono and stereo panoramas using a single image stitching pipeline where the panorama generated by our proposed system is automatically straightened without visible seams. Furthermore, we compared our proposed system with existing mono and stereo contents generation systems in both qualitative and quantitative perspectives, and the comparative measurements obtained verified the effectiveness of our system compared to existing mono and stereo generation systems.