On-device Real-time Custom Hand Gesture Recognition

TL;DR

This paper introduces a user-friendly, on-device framework for real-time custom hand gesture recognition that allows users to easily train and deploy personalized gesture models with minimal ML expertise and no-code tools.

Contribution

It presents a novel on-device system enabling users to customize hand gesture recognition through fine-tuning a pre-trained model with minimal data and no-code deployment options.

Findings

Supports real-time on-device gesture recognition

Allows quick customization with few images per gesture

Accessible to users with limited ML knowledge

Abstract

Most existing hand gesture recognition (HGR) systems are limited to a predefined set of gestures. However, users and developers often want to recognize new, unseen gestures. This is challenging due to the vast diversity of all plausible hand shapes, e.g. it is impossible for developers to include all hand gestures in a predefined list. In this paper, we present a user-friendly framework that lets users easily customize and deploy their own gesture recognition pipeline. Our framework provides a pre-trained single-hand embedding model that can be fine-tuned for custom gesture recognition. Users can perform gestures in front of a webcam to collect a small amount of images per gesture. We also offer a low-code solution to train and deploy the custom gesture recognition model. This makes it easy for users with limited ML expertise to use our framework. We further provide a no-code web front-end for users without any ML expertise. This makes it even easier to build and test the end-to-end pipeline. The resulting custom HGR is then ready to be run on-device for real-time scenarios. This can be done by calling a simple function in our open-sourced model inference API, MediaPipe Tasks. This entire process only takes a few minutes.