OneTrack-M: A multitask approach to transformer-based MOT models

TL;DR

OneTrack-M is a transformer-based multi-object tracking model that simplifies architecture by using only an encoder, achieving at least 25% faster inference while maintaining or improving accuracy, suitable for real-time applications.

Contribution

This work introduces a simplified transformer-based MOT model that eliminates the decoder, reducing processing time and enhancing efficiency through innovative data pre-processing and multitask training.

Findings

Achieves at least 25% faster inference times than state-of-the-art models.

Maintains or improves tracking accuracy metrics.

Suitable for real-time applications like autonomous vehicles and surveillance.

Abstract

Multi-Object Tracking (MOT) is a critical problem in computer vision, essential for understanding how objects move and interact in videos. This field faces significant challenges such as occlusions and complex environmental dynamics, impacting model accuracy and efficiency. While traditional approaches have relied on Convolutional Neural Networks (CNNs), introducing transformers has brought substantial advancements. This work introduces OneTrack-M, a transformer-based MOT model designed to enhance tracking computational efficiency and accuracy. Our approach simplifies the typical transformer-based architecture by eliminating the need for a decoder model for object detection and tracking. Instead, the encoder alone serves as the backbone for temporal data interpretation, significantly reducing processing time and increasing inference speed. Additionally, we employ innovative data pre-processing and multitask training techniques to address occlusion and diverse objective challenges within a single set of weights. Experimental results demonstrate that OneTrack-M achieves at least 25% faster inference times compared to state-of-the-art models in the literature while maintaining or improving tracking accuracy metrics. These improvements highlight the potential of the proposed solution for real-time applications such as autonomous vehicles, surveillance systems, and robotics, where rapid responses are crucial for system effectiveness.