End-to-end Trainable Deep Neural Network for Robotic Grasp Detection and Semantic Segmentation from RGB.
Stefan AinetterFriedrich FraundorferPublished in: ICRA (2021)
Keyphrases
- end to end
- semantic segmentation
- object detection
- text localization and recognition
- label transfer
- superpixels
- weakly supervised
- object categories
- pascal voc
- scene classification
- conditional random fields
- object class
- congestion control
- real time
- viewpoint
- object segmentation
- pixel level
- object classification
- natural scenes
- segmentation method
- color space
- color images
- object recognition
- feature extraction