Fig. 5From: Hands-on: deformable pose and motion models for spatiotemporal localization of fine-grained dyadic interactionsExample frame of a handshake interaction (a). The Platt response space appears in (c). The localization tube is shown in (b), projected on the epitome frame. a Video frame with ground truth box (light blue). b Final detection tube. c Platt response space with the ground truth bounding volume (dashed)Back to article page