# A multimodal spatiotemporal convolutional network with attention mechanism for athlete anxiety behavior recognition

## TL;DR

This paper introduces a system that automatically detects athlete anxiety using combined data from body signals, facial expressions, and movements, offering real-time insights.

## Contribution

A novel multimodal spatiotemporal convolutional network with adaptive attention for real-time athlete anxiety recognition is introduced.

## Key findings

- The system achieves 94.6% accuracy in detecting anxiety in athletes.
- Multimodal data fusion outperforms single-modal methods in anxiety recognition.
- The system supports real-time processing for practical sports applications.

## Abstract

Athletic performance is significantly impacted by anxiety, yet traditional assessment methods rely on subjective questionnaires that lack real-time capability. This study presents an automated anxiety recognition system for athletes using multimodal data fusion of physiological signals, facial expressions, and body movements. The proposed approach employs spatiotemporal convolutional networks with adaptive attention mechanisms to capture behavioral patterns across multiple modalities simultaneously. The system achieves 94.6% accuracy in anxiety detection while maintaining real-time processing capability for practical sports applications. This objective assessment tool enables coaches and sports psychologists to implement timely interventions, potentially improving both athletic performance and athlete mental well-being. The multimodal approach demonstrates significant advantages over single-modal methods, providing a comprehensive solution for anxiety monitoring in competitive sports environments.