Traffic Accident Detection via Self-Supervised Consistency Learning in Driving Scenarios

Abstract

Traffic accident detection and anticipation is an obstinate road safety problem and painstaking efforts have been devoted. With the rapid growth of video data, Vision-based Traffic Accident Detection and Anticipation (named Vision-TAD and Vision-TAA) become the last one-mile problem for safe driving and surveillance safety. However, the long-tailed, unbalanced, highly dynamic, complex, and uncertain properties of traffic accidents form the Out-of-Distribution (OOD) feature for Vision-TAD and Vision-TAA. Current AI development may focus on these OOD but important problems. What has been done for Vision-TAD and Vision-TAA? What direction we should focus on in the future for this problem? A comprehensive survey is important. We present the first survey on Vision-TAD in the deep learning era and the first-ever survey for Vision-TAA. The pros and cons of each research prototype are discussed in detail during the investigation. In addition, we also provide a critical review of 31 publicly available benchmarks and related evaluation metrics. Through this survey, we want to spawn new insights and open possible trends for Vision-TAD and Vision-TAA tasks.

Anomaly Detection Methods in Autonomous Robotic Missions

“… Methods to find anomalies in spatial, temporal and Spatio-temporal elements. Where S.1 [ 33 ], S.2 [ 34 ], S.3 [ 61 ], S.4 [ 35 ], S.5 [ 62 ], S.6 [ 26 ] represent spatial anomaly detection methods, ST.1 [ 42 ], ST.2 [ 43 ], ST.3 [ 44 ], ST.4 [ 46 ], ST.5 [ 47 ], ST.6 [ 45 ] represent Spatio-temporal anomaly detection methods and T.1 [ 38 ], T.2 [ 39 ], T.3 [ 40 ], T.4 [ 41 ] represent Temporal anomaly detection methods. The colour variation represents the year when the method was first used in Robotics for anomaly detection.…”

“…Recent research in robotics has explored spatiotemporal correlations in various contexts. Fang et al [ 45 ] used a graph neural network to capture spatiotemporal correlations in a scene to predict possible accidents during self-driving car missions. Another study on assistive robots in healthcare [ 43 ] initially detected spatial features such as objects in the environment.…”

Collision Risk in Autonomous Vehicles: Classification, Challenges, and Open Research Areas

“…Accordingly, the detection of accidents in traffic flows [48] and identification of dangerous locations on the road using smartphone data [49] were investigated. To detect accidents, one approach is to use multitask adaptation [48].…”

An Attention-guided Multistream Feature Fusion Network for Localization of Risky Objects in Driving Videos

“…Unlike video surveillance systems, dashcam videos capture moving traffic agents that not only rapidly move but appear and disappear quickly in the scene. Different advanced methods are developed to learn the spatiotemporal pattern of the traffic agents to provide an overall riskiness score of the scene, including an LSTM predictor [6], reinforced learning [8], Graph Neural Network [7], [11], and dynamic attention [9]. Although they only predict a risky event in the temporal domain, these studies have developed a solid methodological foundation for risky object localization in the spatial domain.…”

“…Many computer vision studies have tackled a related task that detects anomalous events from a dashcam [6]- [11]. That is, this stream of literature focuses on identifying frames where the risk of a traffic accident is present.…”