Media Summary: Presentation of Stitchable Neural Networks, QPGesture: Quantization-Based and Phase-Guided Motion Matching for Natural Speech-Driven Gesture Generation. [CVPR 2023 Highlight] Autoregressive Visual Tracking
Cvpr 2023 Highlight Gapartnet - Detailed Analysis & Overview
Presentation of Stitchable Neural Networks, QPGesture: Quantization-Based and Phase-Guided Motion Matching for Natural Speech-Driven Gesture Generation. [CVPR 2023 Highlight] Autoregressive Visual Tracking OmniObject3D: Large-Vocabulary 3D Object Dataset for Realistic Perception, Reconstruction and Generation project page: ... A 8-min presentation of our work, "Egocentric Video Task Translation", published at This is the 8 min presentation for MetaPortrait: Identity-preserving Talking Head Generation with Fast Personalized Adaptation.
GOR-IS presents a 3D Gaussian object removal framework that edits scenes in the intrinsic space, enabling physically consistent ... OMG-Bench: A New Challenging Benchmark for Skeleton-based Online Micro Hand Gesture Recognition (