Media Summary: Using Berkeley AUTOLAB Dex-net and GQ-CNN. Use Mask RCNN to segment out the lego blocks, then use DexNet to plan Object grasping using deep learning and point cloud
Using Berkeley AUTOLAB Dex-net and GQ-CNN. Use Mask RCNN to segment out the lego blocks, then use DexNet to plan Object grasping using deep learning and point cloud
Using Berkeley AUTOLAB Dex-net and GQ-CNN.
Near Demo Version.
HSR Mask RCNN Dex-net Grasp Demo
![[HSR-ROBOT] Model-based grasping experiment](https://i.ytimg.com/vi/7fAMVreVPu0/mqdefault.jpg)
Grasping
![[HSR-ROBOT] First grasping experiments with Sasha](https://i.ytimg.com/vi/to7-sqDlNcU/mqdefault.jpg)
First
Use Mask RCNN to segment out the lego blocks, then use DexNet to plan
Object grasping using deep learning and point cloud
HSR
Robot grasping demo