Abstract:Surveillance video has been widely used in business, security, search, and other fields. Identifying and locating specific pedestrians in surveillance video has an important application value in criminal investigation, search and rescue, etc. However, the requirements for real-time capturing and accuracy are high for these applications. It is essential to build a complete and smooth system to combine pedestrian detection, tracking and re-identification to achieve the goal of maximizing efficiency by balancing real-time capture and accuracy. This paper combined the detector and Re-ID models into a single end-to-end network by introducing a new track branch to YOLOv5 architecture for tracking. For pedestrian detection, we employed the weighted bi-directional feature pyramid network (BiFPN) to enhance the network structure based on the YOLOv5-Lite, which is able to further improve the ability of feature extraction. For tracking, based on Deepsort, this paper enhanced the tracker, which uses the Noise Scale Adaptive (NSA) Kalman filter to track, and adds adaptive noise to strengthen the anti-interference of the tracking model. In addition, the matching strategy is further updated. For pedestrian re-identification, the network structure of Fastreid was modified, which can increase the feature extraction speed of the improved algorithm by leaps and bounds. Using the proposed unified network, the parameters of the entire model can be trained in an end-to-end method with the multi-loss function, which has been demonstrated to be quite valuable in some other recent works. Experimental results demonstrate that pedestrians detection can obtain a 97% mean Average Precision (mAP) and that it can track the pedestrians well with a 98.3% MOTA and a 99.8% MOTP on the MOT16 dataset; furthermore, high pedestrian re-identification performance can be achieved on the VERI-Wild dataset with a 77.3% mAP. The overall framework proposed in this paper has remarkable performance in terms of the precise localization and real-time detection of specific pedestrians across time, regions, and cameras.

An RNN-based IMM Filter Surrogate

Handling Missing Observations with an RNN-based Prediction-Update Cycle

IMM Algorithm Based on H∞ Filter for Maneuvering Target Tracking

Interacting multiple model algorithm based on H∞ filter

Imm Algorithm Based On H-Infinity Filter For Maneuvering Target Tracking

Convolutional Recurrent Predictor: Implicit Representation for Multi-target Filtering and Tracking

Intentions of Vulnerable Road Users - Detection and Forecasting by Means of Machine Learning

Long-Term Pedestrian Trajectory Prediction Using Mutable Intention Filter and Warp LSTM

Interacting multiple model particle filter to adaptive visual tracking.

Understanding Pedestrian-Vehicle Interactions with Vehicle Mounted Vision: An LSTM Model and Empirical Analysis

Bimodal Extended Kalman Filter-Based Pedestrian Trajectory Prediction

A Radar-Nearest-Neighbor Based Data-Driven Approach for Crowd Simulation

A Non-linear Pedestrian Tracker Using Velocity-Adaptive Particle Filter with Trajectory Analysis

An Unscented Particle Filter for Ground Maneuvering Target Tracking

Interactive Multiple-Model Learning Filter for Spacecraft Pursuit–Evasion Game Strategy Switch Based on Long Short-Term Memory Network

Stochastic Sampling Simulation for Pedestrian Trajectory Prediction

Spatio-temporal deep learning framework for pedestrian intention prediction in urban traffic scenes

End-to-End Network for Pedestrian Detection, Tracking and Re-Identification in Real-Time Surveillance System

SFMGNet: A Physics-based Neural Network To Predict Pedestrian Trajectories

Novel Interacting Multiple Model Filter for Uncertain Target Tracking Systems Based on Weighted Kullback–Leibler Divergence

Roadside Pedestrian Motion Prediction Using Bayesian Methods and Particle Filter