Learning Deep and Compact Models for Gesture Recognition

Autor:	Mullick, Koustav, Namboodiri, Anoop M.
Rok vydání:	2017
Předmět:	Computer Science - Computer Vision and Pattern Recognition
Druh dokumentu:	Working Paper
Popis:	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. Comment: Accepted at 2017 IEEE International Conference on Image Processing (ICIP 2017)
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/1712.10136 Zobrazit plný text záznamu View this record from Arxiv