论文信息 - CNN Model Design of Gesture Recognition Based on Tensorflow Framework

CNN Model Design of Gesture Recognition Based on Tensorflow Framework

The paper uses Google newest open-source Tensorflow framework to build the model of gesture recognition, introduces the platform characteristics of Tensorflow, and puts forward a convolution network model based on Tensorflow framework. The experiment is designed with the combination of recognized dataset and self-collected dataset. The experimental results show that the model has high recognition accuracy, high computational efficiency, strong robustness, and can easily adjust the network structure, find the optimal model quickly, and accomplish the task of gesture recognition better.

[1] Tomoichi Takahashi,et al. Hand gesture coding based on experiments using a hand gesture interface device , 1991, SGCH.

[2] Yun Tian,et al. CUDA-based real-time hand gesture interaction and visualization for CT volume dataset using leap motion , 2016, The Visual Computer.

[3] Dima Damen,et al. Recognizing linked events: Searching the space of feasible explanations , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[4] JongShill Lee,et al. Hand region extraction and gesture recognition from video stream with complex background through entropy analysis , 2004, The 26th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[5] Geoffrey E. Hinton,et al. ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[6] Geoffrey E. Hinton,et al. Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[7] Lei Hang,et al. Survey of convolutional neural network , 2016 .