🤖 AI Summary
Lightweight keypoint detection for Brazilian Sign Language (LIBRAS) isolated-word recognition faces a fundamental trade-off between recognition accuracy and inference speed. Method: This paper proposes an efficient recognition framework based on optimized keypoint subset selection and missing-keypoint compensation. It employs MediaPipe for body keypoint extraction, introduces a discriminability-driven strategy to identify the minimal effective keypoint subset for sign motion representation, and applies spline interpolation to recover keypoints lost due to occlusion or detection failure. Contribution/Results: The method preserves skeletal motion fidelity while significantly reducing computational redundancy. Experiments demonstrate that, compared to state-of-the-art approaches, our method achieves a 1.2% absolute accuracy gain, a 5.3× speedup in inference latency, and a 68% reduction in model parameters—enabling scalable, robust sign language recognition under resource-constrained conditions.
📝 Abstract
This paper investigates the feasibility of using lightweight body landmark detection for the recognition of isolated signs in Brazilian Sign Language (LIBRAS). Although the skeleton-based approach by Alves et al. (2024) enabled substantial improvements in recognition performance, the use of OpenPose for landmark extraction hindered time performance. In a preliminary investigation, we observed that simply replacing OpenPose with the lightweight MediaPipe, while improving processing speed, significantly reduced accuracy. To overcome this limitation, we explored landmark subset selection strategies aimed at optimizing recognition performance. Experimental results showed that a proper landmark subset achieves comparable or superior performance to state-of-the-art methods while reducing processing time by more than 5X compared to Alves et al. (2024). As an additional contribution, we demonstrated that spline-based imputation effectively mitigates missing landmark issues, leading to substantial accuracy gains. These findings highlight that careful landmark selection, combined with simple imputation techniques, enables efficient and accurate isolated sign recognition, paving the way for scalable Sign Language Recognition systems.