Date and Time:  May 10 (Fri.), 2019, 16:00 – 17:00
Place: Meeting Room #306, Building E-3, UEC
Speaker: Zineb ABDERRAHMANE (École Nationale Supérieure d’Hydrauliqur, Blida, Algeria.)
Chair: Prof. Yoichi MIYAWAKI
Title: Haptic recognition of daily-life objects capable of dealing with data scarcity
Abstract: Recognizing surrounding objects is an important skill for the autonomy of robots performing in daily-life. Nowadays robots are equipped with sophisticated sensors imitating human sensing capabilities such as touch. This allowed to integrate information about object texture, compliance and material ensuing from robot-object physical interaction to the recognition . In this thesis, we aim to exploit machine learning techniques to perform haptic recognition of daily life objects. The main challenge faced in this work is the scarcity of haptic training data for all daily-life objects. This is due to the continuously growing number of objects and the effort and time needed by the robot to physically interact with each object for data collection. We solve this problem by developing a haptic recognition framework capable of performing Zero-shot, One-shot and Multi-shot Learning. We extend this framework by integrating vision to enhance robots performance.