CondenseNet: An Efficient DenseNet using Learned Group Convolutions
This addresses the need for efficient neural networks on mobile devices with limited computational resources, representing a novel architectural improvement rather than an incremental change.
The paper tackles the problem of computational efficiency in deep neural networks for mobile devices by introducing CondenseNet, which combines dense connectivity with learned group convolutions to remove superfluous connections, resulting in far more efficient performance than state-of-the-art compact networks like MobileNets and ShuffleNets.
Deep neural networks are increasingly used on mobile devices, where computational resources are limited. In this paper we develop CondenseNet, a novel network architecture with unprecedented efficiency. It combines dense connectivity with a novel module called learned group convolution. The dense connectivity facilitates feature re-use in the network, whereas learned group convolutions remove connections between layers for which this feature re-use is superfluous. At test time, our model can be implemented using standard group convolutions, allowing for efficient computation in practice. Our experiments show that CondenseNets are far more efficient than state-of-the-art compact convolutional networks such as MobileNets and ShuffleNets.