Master Thesis /Semester Project – Enhancing Prosthetic and Robotic Devices with Neuromorphic Spiking Neural Networks for Tactile Object Classification

Prosthetic and robotic devices play a crucial role in aiding individuals with manipulative disabilities, helping them regain lost abilities and improve their daily lives. These devices often rely on control systems driven by various signals, such as optical, voice, or biomedical signals like EEG and EMG. However, a major limitation in current prosthetic devices is the lack of sophisticated tactile feedback, which hinders their ability to make nuanced decisions about the pressure needed when grasping different objects.

This project aims to develop and implement a neuromorphic spiking neural network (SNN) to enhance the tactile perception and object classification capabilities of prosthetic and robotic devices. By leveraging the advantages of SNNs, this project seeks to address the limitations of traditional convolutional neural networks (CNNs) in handling spatio-temporal tactile data and provide a more efficient and biologically plausible approach to tactile sensing and classification.

Recent advancements include the scalable tactile glove (STAG) [1] that can be used to detect tactile information during grasping. In this project, we will use existing tactile datasets, such as those from the STAG system, which include high-resolution tactile information from various objects [2].

The project will be conducted in collaboration with PBL at ETH.

 

Requirements

Programming skills in Python, interest in neuronal systems (biological or artificial).

 

Reference

[1] Sundaram, S., Kellnhofer, P., Li, Y., Zhu, J. Y., Torralba, A., & Matusik, W. (2019). Learning the signatures of the human grasp using a scalable tactile glove. Nature, 569(7758), 698-702.

[2] Wang, X., Geiger, F., Niculescu, V., Magno, M., & Benini, L. (2022). Leveraging tactile sensors for low latency embedded smart hands for prosthetic and robotic applications. IEEE Transactions on Instrumentation and Measurement, 71, 1-14.

 

Contact

Elisa Donati elisa@ini.uzh.ch