Convolutional neural networks (CNNs) have achieved excellent performance improvement in image processing and other machine learning tasks. However, tremendous computation and memory consumption for most classical CNN models pose a great challenge to the deployment in portable and power-limited devices. In this paper, by analyzing the sensitivity of the rank in each layer of the network accuracy, we propose a sensitivity-oriented layer-wise low-rank approximation algorithm. With specific compression and acceleration requirement, the convolutional layer with higher sensitivity keeps more kernels than that with lower sensitivity. In addition, we also demonstrated that global optimization can obtain a better classification performance than layer-wise fine-tuning. The experimental results show that the proposed method can achieve 20% acceleration ratio gaining compared with the traditional rank-reducing methods.When deployed on the VGG Net-16 model, the proposed method can achieve 2.7× compression/acceleration ratio on convolutional layers and 10.9× compression/acceleration ratio on fully connected (FC) layers with 0.05% top-1 accuracy loss and 0.01% top-5 accuracy loss.
To View the Base Paper Abstract Contents
Now it is Your Time to Shine.
Great careers Start Here.
We Guide you to Every Step
Success! You're Awesome
Thank you for filling out your information!
We’ve sent you an email with your Final Year Project PPT file download link at the email address you provided. Please enjoy, and let us know if there’s anything else we can help you with.
To know more details Call 900 31 31 555
The WISEN Team