Diverse Branch Block: Building a Convolution as an Inception-like Unit

• URL: http://arxiv.org/abs/2103.13425v1
• Date: Wed, 24 Mar 2021 18:12:00 GMT
• Title: Diverse Branch Block: Building a Convolution as an Inception-like Unit
• Authors: Xiaohan Ding, Xiangyu Zhang, Jungong Han, Guiguang Ding
• Abstract summary: We propose a universal building block of Convolutional Neural Network (ConvNet) to improve the performance without any inference-time costs. The Diverse Branch Block (DBB) enhances the representational capacity of a single convolution by combining diverse branches of different scales and complexities. After training, a DBB can be equivalently converted into a single conv layer for deployment.
• Score: 123.59890802196797
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: 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.

Related papers

• Detecting and Approximating Redundant Computational Blocks in Neural Networks [25.436785396394804]
  intra-network similarities present new opportunities for designing more efficient neural networks. We introduce a simple metric, Block Redundancy, to detect redundant blocks, and propose Redundant Blocks Approximation (RBA) to approximate redundant blocks. RBA reduces model parameters and time complexity while maintaining good performance.
  arXiv  Detail & Related papers  (2024-10-07T11:35:24Z)
• Dynamic Perceiver for Efficient Visual Recognition [87.08210214417309]
  We propose Dynamic Perceiver (Dyn-Perceiver) to decouple the feature extraction procedure and the early classification task. A feature branch serves to extract image features, while a classification branch processes a latent code assigned for classification tasks. Early exits are placed exclusively within the classification branch, thus eliminating the need for linear separability in low-level features.
  arXiv  Detail & Related papers  (2023-06-20T03:00:22Z)
• Towards Diverse Binary Segmentation via A Simple yet General Gated Network [71.19503376629083]
  We propose a simple yet general gated network (GateNet) to tackle binary segmentation tasks. With the help of multi-level gate units, the valuable context information from the encoder can be selectively transmitted to the decoder. We introduce a "Fold" operation to improve the atrous convolution and form a novel folded atrous convolution.
  arXiv  Detail & Related papers  (2023-03-18T11:26:36Z)
• Deep ensembles in bioimage segmentation [74.01883650587321]
  In this work, we propose an ensemble of convolutional neural networks (CNNs) In ensemble methods, many different models are trained and then used for classification, the ensemble aggregates the outputs of the single classifiers. The proposed ensemble is implemented by combining different backbone networks using the DeepLabV3+ and HarDNet environment.
  arXiv  Detail & Related papers  (2021-12-24T05:54:21Z)
• Dynamic Convolution for 3D Point Cloud Instance Segmentation [146.7971476424351]
  We propose an approach to instance segmentation from 3D point clouds based on dynamic convolution. We gather homogeneous points that have identical semantic categories and close votes for the geometric centroids. The proposed approach is proposal-free, and instead exploits a convolution process that adapts to the spatial and semantic characteristics of each instance.
  arXiv  Detail & Related papers  (2021-07-18T09:05:16Z)
• Dynamic Clone Transformer for Efficient Convolutional Neural Netwoks [0.0]
  We introduce a novel concept termed multi-path fully connected pattern (MPFC) to rethink the interdependencies of topology pattern, accuracy and efficiency for ConvNets. Inspired by MPFC, we propose a dual-branch module named dynamic clone transformer (DCT) where one branch generates multiple replicas from inputs and another branch reforms those clones through a series of difference vectors conditional on inputs itself to produce more variants.
  arXiv  Detail & Related papers  (2021-06-12T13:42:28Z)
• HS-ResNet: Hierarchical-Split Block on Convolutional Neural Network [37.56074820823266]
  Hierarchical-Split Block can be taken as a plug-and-play block to upgrade existing convolutional neural networks. Our network achieves 81.28% top-1 accuracy with competitive latency on ImageNet-1k dataset.
  arXiv  Detail & Related papers  (2020-10-15T09:32:38Z)
• Structured Convolutions for Efficient Neural Network Design [65.36569572213027]
  We tackle model efficiency by exploiting redundancy in the textitimplicit structure of the building blocks of convolutional neural networks. We show how this decomposition can be applied to 2D and 3D kernels as well as the fully-connected layers.
  arXiv  Detail & Related papers  (2020-08-06T04:38:38Z)
• From Discrete to Continuous Convolution Layers [14.41405579899976]
  We propose a generalization of the common Conv-layer, from a discrete layer to a Continuous Convolution layer. CC layers naturally extend Conv-layers by representing the filter as a learned continuous function over sub-pixel coordinates. This allows learnable and principled resizing of feature maps, to any size, dynamically and consistently across scales.
  arXiv  Detail & Related papers  (2020-06-19T13:16:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.