Measuring Generalisation to Unseen Viewpoints, Articulations, Shapes and Objects for 3D Hand Pose Estimation Under Hand-Object Interaction.
Anil ArmaganGuillermo Garcia-HernandoSeungryul BaekShreyas HampaliMahdi RadZhaohui ZhangShipeng XieMingxiu ChenBoshen ZhangFu XiongYang XiaoZhiguo CaoJunsong YuanPengfei RenWeiting HuangHaifeng SunMarek HrúzJakub KanisZdenek KrnoulQingfu WanShile LiLinlin YangDongheui LeeAngela YaoWeiguo ZhouSijia MeiYunhui LiuAdrian SpurrUmar IqbalPavlo MolchanovPhilippe WeinzaepfelRomain BrégierGrégory RogezVincent LepetitTae-Kyun KimPublished in: ECCV (23) (2020)
Keyphrases
- d objects
- real objects
- visual objects
- object model
- multiple objects
- object shapes
- rotationally symmetric
- viewpoint
- data objects
- three dimensional objects
- object shape
- recognizing objects
- shape descriptors
- human hand
- hand pose
- object contours
- real world objects
- individual objects
- similar objects
- spatial relations
- object level
- complex scenes
- moving objects
- object classes
- object features
- transparent objects
- object class
- line drawings
- object models
- bounding box
- digital objects
- object representation
- object parts
- geometric models
- previously unseen
- target object
- object segmentation
- spatial relationships
- polyhedral objects
- complex shapes
- shape representation
- complex objects
- three dimensional
- human body
- physical objects
- rigid transformation
- pose estimation
- multiple viewpoints
- viewing direction
- range data
- surface patches
- lighting conditions
- visual hull
- multiple views
- shape analysis
- range images
- closed curves
- position and orientation
- object manipulation
- human computer interaction
- object detection
- training set
- object recognition