Pedestrian trajectory prediction method based on the Social-LSTM model for vehicle collision

The coronavirus disease (COVID-19) is primarily disseminated through physical contact. As a precaution, it is recommended that indoor spaces have a limited number of people and at least one meter apart. This study proposes a real-time method for monitoring physical distancing compliance in indoor spaces using computer vision and deep learning techniques. The proposed method utilizes YOLO (You Only Look Once), a popular convolutional neural network-based object detection model, pre-trained on the Microsoft COCO (Common Objects in Context) dataset to detect persons and estimate their physical distance in real time. The effectiveness of the proposed method was assessed using metrics including accuracy rate, frame per second (FPS), and mean average precision (mAP). The results show that the YOLO v3 model had the most remarkable accuracy (87.07%) and mAP (89.91%). On the other hand, the highest fps rate of up to 18.71 was achieved by the YOLO v5s model. The results demonstrate the potential of the proposed method for effectively monitoring physical distancing compliance in indoor spaces, providing valuable insights for future use in other public health scenarios.


 
 
 Recent research on dense captioning based on the recurrent neural network and the convolutional neural network has made a great progress. However, mapping from an image feature space to a description space is a nonlinear and multimodel task, which makes it difficult for the current methods to get accurate results. In this paper, we put forward a novel approach for dense captioning based on hourglass-structured residual learning. Discriminant feature maps are obtained by incorporating dense connected networks and residual learning in our model. Finally, the performance of the approach on the Visual Genome V1.0 dataset and the region labelled MS-COCO (Microsoft Common Objects in Context) dataset are demonstrated. The experimental results have shown that our approach outperforms most current methods.
 
 

Caption generation requires best of both Computer Vision and Natural Language Processing. Due to recent improvements in both of them many efficient models have been developed. Automatic Image Captioning can be utilized to provide descriptions of website content or to engender frame-by-frame descriptions of video for the vision-impaired and in many such applications. In this work, a model is described which is utilized to generate novel image captions for a previously unseen image by utilizing a multimodal architecture by amalgamation of a Recurrent Neural Network (RNN) and a Convolutional Neural Network (CNN). The model is trained on Microsoft Common Objects in Context (MSCOCO), an image captioning dataset that aligns captions and images in the same representation space, so that an image is close to its relevant captions in that space and far away from dissimilar captions and dissimilar images. ResNet-50 architecture is used for extracting features from the images and GloVe embeddings are used along with Gated Recurrent Unit (GRU) in Recurrent Neural Network (RNN) for text representation. MSCOCO evaluation server is used for evaluation of the machine generated caption for a given image.

Automated vehicle detection within the advanced application framework of autonomous vehicles significantly enhances road safety compared to human drivers on roads and highways. However, the intelligent identification of road vehicles in a densely populated country like Bangladesh is challenging due to irregular traffic patterns, highly diverse vehicle types, a cluttered environment, and a lack of high‐quality datasets. This study proposes a system that utilizes computer vision technology to identify road vehicles with greater speed and accuracy. First, the dataset was collected and organized in Roboflow to identify the 21 classes of Bangladeshi native vehicle images, along with two additional classes for people and animals. Subsequently, the You Only Look Once v5 (YOLOv5) model underwent training on the dataset. This process produced bounding boxes, which were then refined using the non‐maximum suppression technique. The loss function complete intersection over union is employed to obtain the accurate regression bounding box of the vehicles. The MS COCO (Microsoft Common Objects in Context) dataset weights are included in the YOLOv5 deep learning network for transfer learning. Finally, Python TensorBoard was used to evaluate and visualize the model's performance. The model was developed and validated on the Google Colab platform. A set of experimental evaluations demonstrate that the proposed method is effective and efficient in recognizing Bangladeshi vehicles. In all test road scenarios, the proposed computer vision system for road vehicle identification achieved 95.8% accuracy and 0.3 ms processing time for 200 epochs. This research could lead to intelligent transportation systems and driverless vehicles in Bangladesh.

Text-to-image synthesis: Starting composite from the foreground content

This article proposes to use computer vision technology to detect the Sun in the mode of unstabilized motion of a nanosatellite, which has not previously been used for these purposes. The proposed technology for detecting objects of a given class is based on the use of a Convolutional Neural Network (CNN) with the You Only Look Once version 3 (YOLOv3) architecture. This version of the CNN architecture makes it easy to organize training using data on selected solar images presented in the Microsoft Common Objects in Context (MS COCO) format. Software has been created that allows analyzing data provided by a neural network, such as time-referenced frames of solar images, visualizing the trajectories of the sun, finding estimates of the angular velocity of a nanosatellite, etc. The conducted simulation confirmed the expediency of using computer vision technology on spacecraft for observing the Sun. At the same time, the architecture of YOLOv3 convolutional neural networks does not impose high requirements on the implementation of CNN on onboard computing facilities and can be implemented on nanosatellites.

Object detection has been a key task in computer vision with deep convolutional neural networks being a significant performer. We propose a method named Region Average Pooling that leverages object co-occurrence to improve object detection performance. Given regions of interest in an image, our method augments object detection networks with pooled contextual features from other regions of interest in the scene. We implement our scheme and evaluate it on the Pascal Visual Object Classes (VOC) 2007 and Microsoft Common Objects in Context (MS COCO) datasets. When used as part of the Faster R-CNN object detection framework with VGG-16, we show an increase in mAP from 24.2% to 25.5% over baseline Faster R-CNN and Global Average Pooling when testing on MS COCO.

Pedestrian trajectory prediction is fundamental to a wide range of scientific research work and industrial applications. Most of the current advanced trajectory prediction methods incorporate context information such as pedestrian neighbourhood, labelled static obstacles, and the background scene into the trajectory prediction process. In contrast to these methods which require rich contexts, the method in our paper focuses on predicting a pedestrian's future trajectory using his/her observed part of the trajectory only. Our method, which we refer to as LVTA, is a Location-Velocity-Temporal Attention LSTM model where two temporal attention mechanisms are applied to the hidden state vectors from the location and velocity LSTM layers. In addition, a location-velocity attention layer embedded inside a tweak module is used to improve the predicted location and velocity coordinates before they are passed to the next time step. Extensive experiments conducted on three large benchmark datasets and comparison with eleven existing trajectory prediction methods demonstrate that LVTA achieves competitive prediction performance. Specifically, LVTA attains 9.19 pixels Average Displacement Error (ADE) and 17.28 pixels Final Displacement Error (FDE) for the Central Station dataset, and 0.46 metres ADE and 0.92 metres FDE for the ETH&UCY datasets. Furthermore, evaluation on using LVTA to generate trajectories of different prediction lengths and on new scenes without the need of retraining confirms that it has good generalizability.

Advancing pedestrian trajectory prediction with interaction-aware 3D-dual contextualized modeling

Non-probability sampling network based on anomaly pedestrian trajectory discrimination for pedestrian trajectory prediction

This paper presents a novel framework for accurate pedestrian trajectory prediction in intersection corners or near crosswalks. Given prior knowledge of curbside geometry (i.e. angle made by intersecting curbs at the corner point of interest and the coordinates of the corner itself), the presented framework can accurately predict pedestrian trajectories even in new, unseen intersections. This is achieved by learning motion primitives in a common frame, called the curbside coordinate frame. A key insight in developing this common frame is to ensure that trajectories from intersections with different geometries, representing the same behavior, are spatially similar in the common frame. Motion primitives learned in such a common frame, can then be easily generalized to predict in new intersections, with different geometries than the ones trained on. We test our algorithm on real pedestrian trajectory datasets collected at two intersections, with distinctly different curbside and crosswalk geometries. A comparison of our algorithm with [1] demonstrates improved prediction accuracies of pedestrian trajectory prediction in the case of same training and test intersections, and the improvement of accuracy in the most different training and test intersections scenarios. The result also shows additional context, such as information about pedestrian traffic lights, if available, can be easily incorporated in our prediction model for further improvement in prediction accuracy.

In this paper, a system solution is designed to improve the service efficiency of intersection. For the purpose, a method of trajectory prediction for vehicle is put forward, which is based on the basis of realize multi-targets tracking. The characteristics of traffic video, structural of road profile and movement of vehicles have been analyzed. We designed the multi-targets tracking algorithm based on particle filter for traffic video sequence, in order to improve the accuracy and rapidity, the heuristic method for multi-targets data correlation is proposed. Based on the results of tracking, we analyze the rule of movement parameters of vehicle as the path has changed, according to the features of structured road the vehicle running status parameters is simplified, and a Hidden Markov Model (HMM) has trained, which is used for vehicle trajectory prediction. In the end, Analysis the basic reason of traffic conflict and the dynamic model of traffic conflict zone is designed. Experimental results demonstrate that, the method of tracking based on particle filter has good tracking performance in traffic video; the vehicle trajectory can be predicted rapidly in the required time, and traffic conflict zone is produced accurately.

Physics constrained pedestrian trajectory prediction with probability quantification

This article introduces a novel model for low-quality pedestrian trajectory prediction, the social nonstationary transformers (NSTransformers), that merges the strengths of NSTransformers and spatiotemporal graph transformer (STAR). The model can capture social interaction cues among pedestrians and integrate features across spatial and temporal dimensions to enhance the precision and resilience of trajectory predictions. We also propose an enhanced loss function that combines diversity loss with logarithmic root mean squared error (log-RMSE) to guarantee the reasonableness and diversity of the generated trajectories. This design adapts well to complex pedestrian interaction scenarios, thereby improving the reliability and accuracy of trajectory prediction. Furthermore, we integrate a generative adversarial network (GAN) to model the randomness inherent in pedestrian trajectories. Compared to the conventional standard Gaussian distribution, our GAN approach better simulates the intricate distribution found in pedestrian trajectories, enhancing the trajectory prediction's diversity and robustness. Experimental results reveal that our model outperforms several state-of-the-art methods. This research opens the avenue for future exploration in low-quality pedestrian trajectory prediction.

Trajectory prediction is a prerequisite for missile and high-speed vehicle guidance interception. To address the trajectory problem for the near space hypersonic unpowered gliding vehicle. Firstly, the trajectory prediction mechanism is analyzed. Based on the existing NSHV trajectory prediction method, the key techniques of trajectory prediction are discussed from the perspectives of prediction parameter selection and prediction parameter description. Then, a trajectory prediction method based on Hough transform is proposed. The core of the method is to select the longitude latitude, and high direction positions as prediction parameters, and use the Hough transform to fit the prior basis functions. In the latitude and longitude direction, the rationality of predicting parameters is discussed from two aspects: position extreme point and lateral maneuverability. In the height direction, for the NSHV oscillation drop characteristic, two Hough transforms are used to separate the periodic term from the linear term. In addition, we give the approximate interval of the parameter values of the method, and design a parameter adaptive scheme. Finally, trajectory prediction is performed on two different control patterns. The error of the prediction time of 100s is within 25km, and the results show the effectiveness of the proposed method.