Diverse Branch Block: Building a Convolution as an Inception-like Unit
This addresses the need for more efficient and powerful ConvNets in computer vision tasks like image classification, object detection, and semantic segmentation, though it is incremental as it builds on existing architectures.
The paper tackles the problem of improving Convolutional Neural Network performance without increasing inference-time costs by proposing Diverse Branch Block (DBB), which enhances representational capacity through diverse branches and can be converted to a single convolution layer after training, achieving up to 1.9% higher top-1 accuracy on ImageNet.
We propose a universal building block of Convolutional Neural Network (ConvNet) to improve the performance without any inference-time costs. The block is named Diverse Branch Block (DBB), which enhances the representational capacity of a single convolution by combining diverse branches of different scales and complexities to enrich the feature space, including sequences of convolutions, multi-scale convolutions, and average pooling. After training, a DBB can be equivalently converted into a single conv layer for deployment. Unlike the advancements of novel ConvNet architectures, DBB complicates the training-time microstructure while maintaining the macro architecture, so that it can be used as a drop-in replacement for regular conv layers of any architecture. In this way, the model can be trained to reach a higher level of performance and then transformed into the original inference-time structure for inference. DBB improves ConvNets on image classification (up to 1.9% higher top-1 accuracy on ImageNet), object detection and semantic segmentation. The PyTorch code and models are released at https://github.com/DingXiaoH/DiverseBranchBlock.