VarGNet: Variable Group Convolutional Neural Network for Efficient Embedded Computing

In this paper, we propose a novel network design mechanism for efficient embedded computing. Inspired by the limited computing patterns, we propose to fix the number of channels in a group convolution, instead of the existing practice that fixing the total group numbers. Our solution based network, named Variable Group Convolutional Network (VarGNet), can be optimized easier on hardware side, due to the more unified computing schemes among the layers. Extensive experiments on various vision tasks, including classification, detection, pixel-wise parsing and face recognition, have demonstrated the practical value of our VarGNet.

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Face Verification AgeDB-30 VarGNet Accuracy 0.97333 # 4
Face Verification CFP-FP VarGNet Accuracy 0.89829 # 4
Face Verification Labeled Faces in the Wild VarGNet Accuracy 99.733% # 5

Methods


No methods listed for this paper. Add relevant methods here