Learning Visual and Motion Aware 3D Model Representations for Semantic Retrieval.
Virtual Reality (VR) content creators today face lengthy trial-and-error cycles when searching for 3D models, which relies heavily on manually annotated metadata. However, this approach struggles to capture the rich visual and dynamic details inherent to 3D models. Prior work has attempted to address this by rendering...