A Multi-task Learning Framework for Grasping-Position Detection and Few-Shot Classification

no code implementations • 12 Mar 2020 • Yasuto Yokota, Kanata Suzuki, Yuzi Kanazawa, Tomoyoshi Takebayashi

However, the DNN that was used to detect grasping positions has two problems with respect to extracting feature vectors from a layer for shape classification: (1) Because each layer of the grasping position detection DNN is activated by all objects in the input image, it is necessary to refine the features for each grasping position.

Robotics

Online Self-Supervised Learning for Object Picking: Detecting Optimum Grasping Position using a Metric Learning Approach

no code implementations • 8 Mar 2020 • Kanata Suzuki, Yasuto Yokota, Yuzi Kanazawa, Tomoyoshi Takebayashi

: SSD that detects the grasping position of an object, and Siamese networks (SNs) that evaluate the trial sample using the similarity of two input data in the feature space.

Metric Learning Object +2