UIST 2024 ：EMS Hand Prop

My Contributions :

1. A novel concept of leveraging the loss of SoA caused by EMS to make the stimulated hand serve better as a virtual object

2. Proposal of a highly reproducible electrode layout with a clear anatomical guide for actuating fingers via EMS and inducing users to pose 8 gestures with it

3. An interaction system that combined EMS actuation, deep learning model, and data glove to enable users to perform gestural object retrieval tasks and be involved in interactive scenarios easily and immersively

This figure shows the electrode layout we provided, where each finger has an electrode attached to it, responsible for actuating that finger. This electrode layout is based on BackHandEMS, which misses an explicit anatomical guide for electrode layout to stimulate interossei and lumbrical. Ours provides a highly reproducible electrode layout with a clear anatomical guide of the muscles for flexing each single finger.

This figure shows 8 examples of actuating gestures via EMS and the virtual object corresponding to the gesture. Gestures design can be referred to HandInterfaces.

This figure shows how to recognize these gestures using the data from data glove and simple neural network.

Here is a demo video of the system:

Unfortunately, our paper is rejected after rebuttal. Currently, we are submitting it to Internation Journal of Human-Computer Studies