Learning Deep and Compact Models for Gesture Recognition

2017/12/29 by Mullick, Koustav, Namboodiri, Anoop M. · 1 citation
#Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.1712.10136

Abstract

We look at the problem of developing a compact and accurate model for gesture recognition from videos in a deep-learning framework. Towards this we propose a joint 3DCNN-LSTM model that is end-to-end trainable and is shown to be better suited to capture the dynamic information in actions. The solution achieves close to state-of-the-art accuracy on the ChaLearn dataset, with only half the model size. We also explore ways to derive a much more compact representation in a knowledge distillation framework followed by model compression. The final model is less than 1~MB in size, which is less than one hundredth of our initial model, with a drop of 7% in accuracy, and is suitable for real-time gesture recognition on mobile devices.

Cited by

• How do Foundation Models Compare to Skeleton-Based Approaches for Gesture Recognition in Human-Robot Interaction?

Related

• DeepGRU: Deep Gesture Recognition Utility
• Fast Learning of Dynamic Hand Gesture Recognition with Few-Shot Learning Models
• Large-scale Isolated Gesture Recognition Using Convolutional Neural Networks
• Learning to recognize touch gestures: recurrent vs. convolutional features and dynamic sampling
• Large-scale Continuous Gesture Recognition Using Convolutional Neural Networks