Multi-Residual Networks: Improving the Speed and Accuracy of Residual Networks
This work addresses the need for more efficient and accurate convolutional neural networks for image classification tasks, representing an incremental improvement over existing residual network architectures.
The authors tackled the problem of improving the speed and accuracy of residual networks by proposing a multi-residual network architecture that increases residual functions in blocks, making models wider rather than deeper. This achieved error rates of 3.73% on CIFAR-10 and 19.45% on CIFAR-100, outperforming most existing models, and improved top-1 error by 0.22% on ImageNet 2012, with a model parallelism technique reducing computational complexity by up to 15%.
In this article, we take one step toward understanding the learning behavior of deep residual networks, and supporting the observation that deep residual networks behave like ensembles. We propose a new convolutional neural network architecture which builds upon the success of residual networks by explicitly exploiting the interpretation of very deep networks as an ensemble. The proposed multi-residual network increases the number of residual functions in the residual blocks. Our architecture generates models that are wider, rather than deeper, which significantly improves accuracy. We show that our model achieves an error rate of 3.73% and 19.45% on CIFAR-10 and CIFAR-100 respectively, that outperforms almost all of the existing models. We also demonstrate that our model outperforms very deep residual networks by 0.22% (top-1 error) on the full ImageNet 2012 classification dataset. Additionally, inspired by the parallel structure of multi-residual networks, a model parallelism technique has been investigated. The model parallelism method distributes the computation of residual blocks among the processors, yielding up to 15% computational complexity improvement.