ImVoxelNet: Image to Voxels Projection for Monocular and Multi-View General-Purpose 3D Object Detection.
Danila RukhovichAnna VorontsovaAnton KonushinPublished in: WACV (2022)
Keyphrases
- multi view
- d objects
- multi views
- multiple views
- single view
- pose estimation
- camera calibration
- depth map
- input image
- object recognition
- three dimensional
- viewpoint
- single image
- depth cues
- range data
- pose parameters
- image features
- photometric stereo
- bundle adjustment
- light field
- d mesh
- feature points
- face recognition
- range images
- image synthesis
- ego motion
- image sequences
- vanishing points
- high resolution
- camera parameters
- shape descriptors
- image matching
- object classes
- lighting conditions
- machine learning
- semi supervised
- object detection
- optical flow