COPE: End-to-end trainable Constant Runtime Object Pose Estimation.
Stefan ThalhammerTimothy PattenMarkus VinczePublished in: CoRR (2022)
Keyphrases
- end to end
- pose estimation
- d objects
- object recognition and pose estimation
- position and orientation
- object pose
- invariant recognition
- pose recovery
- human pose
- degrees of freedom
- human body
- computer vision
- pose determination
- pose parameters
- feature points
- head pose estimation
- object recognition
- monocular images
- shape from silhouettes
- multipath
- viewpoint
- human pose estimation
- moving objects
- complex objects
- three dimensional
- object model
- tracking and pose estimation
- bounding box
- computational complexity
- implicit polynomial
- body parts
- congestion control
- object segmentation
- camera pose
- visual hull
- multi view
- articulated objects
- camera pose estimation
- machine learning