Enhanced Transformer-Based Tracking for Skiing Events: Overcoming Multi-Camera Challenges, Scale Variations and Rapid Motion -- SkiTB Visual Tracking Challenge 2025

TL;DR

This paper presents an adapted transformer-based model, STARK, for accurate skier tracking in alpine sports, addressing challenges like occlusions, camera movements, and environmental variations to improve performance analysis and injury prevention.

Contribution

We adapt the STARK transformer-based tracking model specifically for skiing events, overcoming domain-specific challenges such as occlusions and camera changes.

Findings

Enhanced tracking accuracy in skiing scenarios

Robustness to occlusions and camera movements

Improved performance over traditional methods

Abstract

Accurate skier tracking is essential for performance analysis, injury prevention, and optimizing training strategies in alpine sports. Traditional tracking methods often struggle with occlusions, dynamic movements, and varying environmental conditions, limiting their effectiveness. In this work, we used STARK (Spatio-Temporal Transformer Network for Visual Tracking), a transformer-based model, to track skiers. We adapted STARK to address domain-specific challenges such as camera movements, camera changes, occlusions, etc. by optimizing the model's architecture and hyperparameters to better suit the dataset.