Viewpoints and Keypoints
arXiv:1411.6067
Abstract
We characterize the problem of pose estimation for rigid objects in terms of determining viewpoint to explain coarse pose and keypoint prediction to capture the finer details. We address both these tasks in two different settings - the constrained setting with known bounding boxes and the more challenging detection setting where the aim is to simultaneously detect and correctly estimate pose of objects. We present Convolutional Neural Network based architectures for these and demonstrate that leveraging viewpoint estimates can substantially improve local appearance based keypoint predictions. In addition to achieving significant improvements over state-of-the-art in the above tasks, we analyze the error modes and effect of object characteristics on performance to guide future efforts towards this goal.
References in corpus (6)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Caffe: Convolutional Architecture for Fast Feature Embedding
- DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition
- DeepPose: Human Pose Estimation via Deep Neural Networks
- Rich feature hierarchies for accurate object detection and semantic segmentation
- Deformable Part Models are Convolutional Neural Networks
Cited by in corpus (16)
- Render for CNN: Viewpoint Estimation in Images Using CNNs Trained with Rendered 3D Model Views
- Recurrent Network Models for Human Dynamics
- PointFusion: Deep Sensor Fusion for 3D Bounding Box Estimation
- Pix3D: Dataset and Methods for Single-Image 3D Shape Modeling
- WarpNet: Weakly Supervised Matching for Single-view Reconstruction
- 3D Bounding Box Estimation Using Deep Learning and Geometry
- Deep Cuboid Detection: Beyond 2D Bounding Boxes
- Crafting a multi-task CNN for viewpoint estimation
- Convolutional Models for Joint Object Categorization and Pose Estimation
- Deep Supervision with Shape Concepts for Occlusion-Aware 3D Object Parsing
- Bottom-Up and Top-Down Reasoning with Hierarchical Rectified Gaussians
- Pose Induction for Novel Object Categories
- Click Here: Human-Localized Keypoints as Guidance for Viewpoint Estimation
- Learning Where to Look: Data-Driven Viewpoint Set Selection for 3D Scenes
- Object Pose Estimation from Monocular Image using Multi-View Keypoint Correspondence
- Learning 3D Deformation of Animals from 2D Images