Women Sport Actions Dataset for Visual Classification Using Small Scale Training Data

2025-07-15

AI Summary

To address the scarcity of female sports action imagery and insufficient modeling of intra-class and inter-class variations—key bottlenecks in few-shot action recognition—this work introduces WomenSports, the first dedicated benchmark dataset for visual classification of women’s sports actions, featuring fine-grained samples across diverse scenes, poses, and attire. Methodologically, we propose a Local Context Region-based Channel Attention (LCRA) mechanism, integrated into ResNet-50 to enhance discriminative feature learning. On WomenSports, our approach achieves 89.15% Top-1 accuracy. Cross-dataset evaluation further demonstrates strong generalization capability, significantly outperforming baseline methods. This work bridges dual gaps in the field: it provides the first large-scale, gender-specific action dataset and a tailored attention architecture, thereby establishing a foundational resource for fair, robust, and inclusive sports motion analysis.

Technology Category

• Computer Vision: Video Understanding & Activity Analysis
• Application Domains: Humanities & Computational Social Science
• Machine Learning: Learning with Manifolds

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Sports action classification representing complex body postures and player-object interactions is an emerging area in image-based sports analysis. Some works have contributed to automated sports action recognition using machine learning techniques over the past decades. However, sufficient image datasets representing women sports actions with enough intra- and inter-class variations are not available to the researchers. To overcome this limitation, this work presents a new dataset named WomenSports for women sports classification using small-scale training data. This dataset includes a variety of sports activities, covering wide variations in movements, environments, and interactions among players. In addition, this study proposes a convolutional neural network (CNN) for deep feature extraction. A channel attention scheme upon local contextual regions is applied to refine and enhance feature representation. The experiments are carried out on three different sports datasets and one dance dataset for generalizing the proposed algorithm, and the performances on these datasets are noteworthy. The deep learning method achieves 89.15% top-1 classification accuracy using ResNet-50 on the proposed WomenSports dataset, which is publicly available for research at Mendeley Data.

Problem

• Lack of diverse women sports datasets for classification
• Need for effective small-scale training data solutions
• Improving feature representation in sports action recognition

Innovation

• Introduces WomenSports dataset for small-scale training
• Uses CNN with channel attention for feature refinement
• Achieves 89.15% accuracy with ResNet-50