Abstract The evaluation of basketball free throw techniques has traditionally relied on subjective assessments, which introduces inherent biases and inconsistencies in performance analysis. This study presents PoseShot, a novel dual channel hybrid CNN-BiLSTM-Transformer Model that facilitates the comprehensive analysis of free throw mechanics with data-driven insights. Unlike conventional human activity recognition that focuses on coarse activity labels, PoseShot is designed to analyze fine-grained, phase-dependent mechanics within a single basketball free throw motion. The proposed framework integrates training footage with precise body posture angle calculations via a dual-channel deep learning architecture to enable the capture of subtle technical variations. Our innovative approach synthesizes convolutional neural networks (CNN) for spatial feature extraction, bidirectional long short-term memory (BiLSTM) for temporal sequence processing, and a transformer encoder for enhanced contextual understanding of motion dynamics. The model demonstrates exceptional performance, achieving an F 1 -score of 95.76%, precision of 95.72%, and recall of 95.80%. These metrics surpass the performance of established architectures, including DenseNet, Swin Transformer, and Vision Transformer, particularly for the analysis of complex throwing motions. By incorporating both spatial features and postural dynamics, PoseShot provides accuracy in motion analysis. Empirical evaluation reveals PoseShot’s capacity to identify crucial biomechanical determinants of successful free throws, thus offering quantifiable insights for performance enhancement. Since the model’s analysis elucidates the intricate relationship between posture optimization and action consistency, it can provide actionable guidance for athletes and coaches. This research bridges the gap between subjective evaluation methods and advanced motion analytics, establishing PoseShot as a transformative tool in sports performance analysis. The findings demonstrate the potential for data-driven approaches to revolutionize basketball training methodologies through precise, objective assessment criteria.
Hsu et al. (Sun,) studied this question.