MDHA: Multi-Scale Deformable Transformer with Hybrid Anchors for Multi-View 3D Object Detection.
Michelle AdelineJunn Yong LooVishnu Monn BaskaranPublished in: CoRR (2024)
Keyphrases
- multi view
- d objects
- multiscale
- multiple views
- single view
- object recognition
- viewpoint
- three dimensional
- camera calibration
- depth map
- object detection
- photometric stereo
- multi views
- range images
- pose estimation
- detection method
- range data
- surface reconstruction
- view synthesis
- visual hull
- shape descriptors
- texture mapping
- articulated objects
- bundle adjustment
- object classes
- computer vision
- keypoints
- control points
- d mesh
- image processing
- semi supervised
- coarse to fine
- human body