ImGeoNet: Image-induced Geometry-aware Voxel Representation for Multi-view 3D Object Detection.
Tao TuShun-Po ChuangYu-Lun LiuCheng SunKe ZhangDonna RoyCheng-Hao KuoMin SunPublished in: CoRR (2023)
Keyphrases
- multi view
- d objects
- multi views
- three dimensional
- multiple views
- single view
- camera calibration
- image features
- geometric models
- depth map
- single image
- pose parameters
- topological information
- input image
- image data
- object recognition
- geometric constraints
- feature points
- d mesh
- viewpoint
- camera parameters
- scene geometry
- photometric stereo
- vanishing points
- surface points
- view synthesis
- light field
- semi supervised
- test images
- bundle adjustment
- shape descriptors
- viewing direction
- complex scenes
- surface reconstruction
- range images
- image synthesis
- object detection
- pose estimation
- machine learning
- image based rendering
- reconstruction method
- multiple objects
- range data