Semi-automatic Data Annotation System for Multi-Target Multi-Camera Vehicle Tracking

TL;DR

This paper introduces a semi-automatic annotation system that accelerates the creation of real-world multi-camera vehicle tracking datasets by combining automatic trajectory extraction with manual cross-camera matching, addressing data scarcity issues.

Contribution

The proposed system integrates automatic single-camera trajectory generation with manual cross-camera matching, improving dataset creation efficiency for real-world MTMCT applications.

Findings

System significantly reduces annotation time.

Achieves high accuracy in trajectory matching.

Facilitates dataset development for real-world scenarios.

Abstract

Multi-target multi-camera tracking (MTMCT) plays an important role in intelligent video analysis, surveillance video retrieval, and other application scenarios. Nowadays, the deep-learning-based MTMCT has been the mainstream and has achieved fascinating improvements regarding tracking accuracy and efficiency. However, according to our investigation, the lacking of datasets focusing on real-world application scenarios limits the further improvements for current learning-based MTMCT models. Specifically, the learning-based MTMCT models training by common datasets usually cannot achieve satisfactory results in real-world application scenarios. Motivated by this, this paper presents a semi-automatic data annotation system to facilitate the real-world MTMCT dataset establishment. The proposed system first employs a deep-learning-based single-camera trajectory generation method to automatically extract trajectories from surveillance videos. Subsequently, the system provides a recommendation list in the following manual cross-camera trajectory matching process. The recommendation list is generated based on side information, including camera location, timestamp relation, and background scene. In the experimental stage, extensive results further demonstrate the efficiency of the proposed system.