Contrastive Learning for Multi-Object Tracking with Transformers

Pierre-François De Plaen, Nicola Marinello, Marc Proesmans, Tinne Tuytelaars, Luc Van Gool
11/14/2023
cs.CV

No official code found. Best implementation: pfdp0/ContrasTR (9 stars, last commit 2024-11-06).

Abstract

The DEtection TRansformer (DETR) opened new possibilities for object detection by modeling it as a translation task: converting image features into object-level representations. Previous works typically add expensive modules to DETR to perform Multi-Object Tracking (MOT), resulting in more complicated architectures. We instead show how DETR can be turned into a MOT model by employing an instance-level contrastive loss, a revised sampling strategy and a lightweight assignment method. Our training scheme learns object appearances while preserving detection capabilities and with little overhead. Its performance surpasses the previous state-of-the-art by +2.6 mMOTA on the challenging BDD100K dataset and is comparable to existing transformer-based methods on the MOT17 dataset.

View on arXivView PDF

Code Implementations(1)

pfdp0/ContrasTRCommunity100%

Official implementation of ContrasTR: Contrastive Learning for Multi-Object Tracking with Transformers

91C++, CudaAug 17, 20241 years agoNOASSERTION

Similar Papers

A Real-Time DETR Approach to Bangladesh Road Object Detection for Autonomous Vehicles

90% match2024

TaskCLIP: Extend Large Vision-Language Model for Task Oriented Object Detection

88% match2024

External Prompt Features Enhanced Parameter-efficient Fine-tuning for Salient Object Detection

88% match2024

Fine-Tuning Florence2 for Enhanced Object Detection in Un-constructed Environments: Vision-Language Model Approach

87% match2025

Continual Detection Transformer for Incremental Object Detection

87% match1 repo1122026

Semantic-Aligned Matching for Enhanced DETR Convergence and Multi-Scale Feature Fusion

87% match1 repo3012022

Cite this paper

@article{plaen2023contrastive,
  title  = {Contrastive Learning for Multi-Object Tracking with Transformers},
  author = {Pierre-François De Plaen and Nicola Marinello and Marc Proesmans and Tinne Tuytelaars and Luc Van Gool},
  year   = {2023},
  eprint = {2311.08043},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2311.08043v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2311.08043.svg)](https://codeofpaper.com/paper/2311.08043)

Discussion