VSD-MOT: End-to-End Multi-Object Tracking in Low-Quality Video Scenes Guided by Visual Semantic Distillation

TL;DR

This paper introduces VSD-MOT, a novel multi-object tracking framework that leverages visual semantic distillation and adaptive weighting to improve tracking accuracy in low-quality videos while maintaining performance in standard scenarios.

Contribution

The paper proposes a new end-to-end multi-object tracking method guided by visual semantic distillation, incorporating a knowledge distillation framework and dynamic weight regulation for low-quality videos.

Findings

Significantly improves tracking accuracy in low-quality videos

Maintains strong performance in conventional video scenarios

Demonstrates effectiveness through extensive experiments

Abstract

Existing multi-object tracking algorithms typically fail to adequately address the issues in low-quality videos, resulting in a significant decline in tracking performance when image quality deteriorates in real-world scenarios. This performance degradation is primarily due to the algorithms' inability to effectively tackle the problems caused by information loss in low-quality images. To address the challenges of low-quality video scenarios, inspired by vision-language models, we propose a multi-object tracking framework guided by visual semantic distillation (VSD-MOT). Specifically, we introduce the CLIP Image Encoder to extract global visual semantic information from images to compensate for the loss of information in low-quality images. However, direct integration can substantially impact the efficiency of the multi-object tracking algorithm. Therefore, this paper proposes to extract visual semantic information from images through knowledge distillation. This method adopts a teacher-student learning framework, with the CLIP Image Encoder serving as the teacher model. To enable the student model to acquire the capability of extracting visual semantic information suitable for multi-object tracking tasks from the teacher model, we have designed the Dual-Constraint Semantic Distillation method (DCSD). Furthermore, to address the dynamic variation of frame quality in low-quality videos, we propose the Dynamic Semantic Weight Regulation (DSWR) module, which adaptively allocates fusion weights based on real-time frame quality assessment. Extensive experiments demonstrate the effectiveness and superiority of the proposed method in low-quality video scenarios in the real world. Meanwhile, our method can maintain good performance in conventional scenarios.