LatentFusion: End-to-End Differentiable Reconstruction and Rendering for Unseen Object Pose Estimation.
Keunhong ParkArsalan MousavianYu XiangDieter FoxPublished in: CoRR (2019)
Keyphrases
- end to end
- pose estimation
- d objects
- shape from silhouettes
- object recognition and pose estimation
- visual hull
- object pose
- position and orientation
- invariant recognition
- pose recovery
- computer vision
- pose parameters
- human pose
- human body
- three dimensional
- head pose estimation
- camera pose estimation
- computer graphics
- body parts
- congestion control
- human pose estimation
- feature points
- multiple cameras
- object tracking
- viewpoint
- implicit surfaces
- training set
- monocular images
- depth images
- articulated model
- object recognition
- focal length
- image based rendering
- range data
- pictorial structure
- ground plane
- hand pose
- single image
- high quality
- d scene
- pose estimates
- camera calibration
- machine learning