Subject-Level Unknown-Identity Identification from Leap Motion Controller 2 Hand Landmarks

2026-06-22

AI Summary

This study addresses open-set hand-based identity verification using only hand keypoint data, with the dual objective of accurately recognizing enrolled users and rejecting unregistered ones. Leveraging data collected via the Leap Motion Controller 2 under a leave-one-subject-out evaluation protocol, the authors propose a compact and interpretable hand descriptor that integrates fingertip-to-palm distances and palm-normalized inter-finger angles to preserve geometric structure. An inner-loop validation mechanism is introduced to provide an unbiased estimate of the rejection threshold. Extensive experiments compare Extra Trees, centroid-matching cosine embeddings, and MLP+OpenMax, demonstrating that Extra Trees achieves superior performance. The results validate that efficient, contactless hand identity verification and rejection can be accomplished using keypoint data alone.

Abstract

This work studies subject recognition from Leap Motion Controller 2 (LMC2) hand landmark data under a subject-level unknown-identity identification protocol on the Multi View Leap2 Hand Pose (ML2HP) dataset. Using only the landmark modality, we retain the original geometric representation and enrich it with fingertip-to-palm distances and palm-normalized inter-finger angular descriptors. Evaluation is performed under a Leave-One-Subject-Out (LOSO) protocol in which, for each outer fold, one subject is excluded from the enrolled set and treated as unknown at test time. To avoid tuning on the true outer unknown subject, the unknown-rejection threshold is selected in an inner validation step by temporarily withholding one enrolled subject from the inner gallery and using it only for threshold estimation. We compare a tree ensemble baseline with two neural alternatives: a learned embedding baseline based on centroid matching and cosine-similarity-based rejection, and an MLP+OpenMax model, which represents a more established open-set recognition approach. Under this evaluation setup, Extra Trees remains the strongest overall method, indicating that the main challenge on this benchmark is not enrolled-subject discrimination alone, but robust score separation between known and unknown probes. The results support the feasibility of compact, interpretable landmark-based descriptors for contactless hand-based unknown-subject rejection and identification on a small-cohort dataset.

Problem

• unknown-identity identification
• hand landmarks
• open-set recognition
• subject-level recognition
• biometric identification

Innovation

• unknown-identity identification
• hand landmarks
• open-set recognition
• Leave-One-Subject-Out
• geometric descriptors