Benchmark:

MOT20 |

Short name:

LITE

Detector:

Public

Description:

The Lightweight Integrated Tracking-Feature Extraction (LITE)
paradigm is introduced as a novel multi-object tracking (MOT) approach. It enhances ReID-based trackers by eliminating inference, preprocessing, post-processing, and ReID model training costs. LITE uses
real-time appearance features without compromising speed. By integrating appearance feature extraction directly into the tracking pipeline using
standard CNN-based detectors such as YOLOv8m, LITE demonstrates
significant performance improvements. The simplest implementation of
LITE on top of classic DeepSORT achieves a HOTA score of 43.03% at
28.3 FPS on the MOT17 benchmark, making it twice as fast as DeepSORT on MOT17 and four times faster on the more crowded MOT20
dataset, while maintaining similar accuracy. Additionally, a new evaluation framework for tracking-by-detection approaches reveals that conventional trackers like DeepSORT remain competitive with modern stateof-the-art trackers when evaluated under fair conditions. The code will
be available post-publication at https://github.com/Jumabek/LITE

Reference:

J. Alikhanov, D. Obidov, H. Kim. LITE: A Paradigm Shift in Multi-Object Tracking with Efficient ReID Feature Integration. In , 2024.

Last submitted:

December 18, 2024 (3 days ago)

Published:

December 18, 2024 at 02:10:56 CET

Submissions:

2

Project page / code:

https://github.com/jumabek/lite

Open source:

Yes

Hardware:

NVIDIA GeForce RTX 3090

Runtime:

149.3 Hz

Benchmark performance:

Detailed performance:

Raw data:

download