Bidirectional Generation of Object Images and Positions using Deep Generative Models for Service Robotics Applications.
Kaede HayashiWenru ZhengLotfi El HafiYoshinobu HagiwaraTadahiro TaniguchiPublished in: SII (2021)
Keyphrases
- natural images
- visual patterns
- input image
- test images
- generative model
- image statistics
- object recognition
- object parts
- discriminative models
- keypoints
- image regions
- d objects
- probabilistic model
- prior knowledge
- image retrieval
- image features
- computer vision
- spatial relations
- partial occlusion
- multiple objects
- object class
- articulated objects
- object localization
- discriminative learning
- deep learning
- conditional random fields
- image matching
- segmentation algorithm
- image classification
- face recognition