HOSNeRF: Dynamic Human-Object-Scene Neural Radiance Fields from a Single Video.
Jia-Wei LiuYan-Pei CaoTianyuan YangEric Zhongcong XuJussi KeppoYing ShanXiaohu QieMike Zheng ShouPublished in: CoRR (2023)
Keyphrases
- video sequences
- video scene
- object motion
- moving objects
- multiple objects
- object detection and tracking
- dynamic textures
- complex scenes
- object tracking
- dynamic scenes
- target object
- static images
- video images
- video data
- stationary camera
- combining information from multiple
- visual data
- photometric properties
- d objects
- lighting conditions
- object models
- human activities
- video frames
- moving camera
- real objects
- single image
- image sequences
- dynamic range
- visual input
- object detection
- visual saliency
- high dynamic range
- d scene
- video streams
- video analysis
- illumination conditions
- computer vision
- input image
- global illumination
- space time
- spatial relations