Data-Driven Neuron Allocation for Scale Aggregation Networks

04/20/2019
by   Yi Li, et al.
14

Successful visual recognition networks benefit from aggregating information spanning from a wide range of scales. Previous research has investigated information fusion of connected layers or multiple branches in a block, seeking to strengthen the power of multi-scale representations. Despite their great successes, existing practices often allocate the neurons for each scale manually, and keep the same ratio in all aggregation blocks of an entire network, rendering suboptimal performance. In this paper, we propose to learn the neuron allocation for aggregating multi-scale information in different building blocks of a deep network. The most informative output neurons in each block are preserved while others are discarded, and thus neurons for multiple scales are competitively and adaptively allocated. Our scale aggregation network (ScaleNet) is constructed by repeating a scale aggregation (SA) block that concatenates feature maps at a wide range of scales. Feature maps for each scale are generated by a stack of downsampling, convolution and upsampling operations. The data-driven neuron allocation and SA block achieve strong representational power at the cost of considerably low computational complexity. The proposed ScaleNet, by replacing all 3x3 convolutions in ResNet with our SA blocks, achieves better performance than ResNet and its outstanding variants like ResNeXt and SE-ResNet, in the same computational complexity. On ImageNet classification, ScaleNets absolutely reduce the top-1 error rate of ResNets by 1.12 (101 layers) and 1.82 (50 layers). On COCO object detection, ScaleNets absolutely improve the mmAP with backbone of ResNets by 3.6 (101 layers) and 4.6 (50 layers) on Faster RCNN, respectively. Code and models are released at https://github.com/Eli-YiLi/ScaleNet.

READ FULL TEXT

page 1

page 7

research
05/18/2018

MDSSD: Multi-scale Deconvolutional Single Shot Detector for small objects

In order to improve the detection accuracy for objects at different scal...
research
07/13/2020

PSConv: Squeezing Feature Pyramid into One Compact Poly-Scale Convolutional Layer

Despite their strong modeling capacities, Convolutional Neural Networks ...
research
11/01/2021

Dense Prediction with Attentive Feature Aggregation

Aggregating information from features across different layers is an esse...
research
07/23/2022

SSBNet: Improving Visual Recognition Efficiency by Adaptive Sampling

Downsampling is widely adopted to achieve a good trade-off between accur...
research
09/25/2021

TreeNet: A lightweight One-Shot Aggregation Convolutional Network

The architecture of deep convolutional networks (CNNs) has evolved for y...
research
04/02/2019

Res2Net: A New Multi-scale Backbone Architecture

Representing features at multiple scales is of great importance for nume...
research
03/24/2021

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

We propose a universal building block of Convolutional Neural Network (C...

Please sign up or login with your details

Forgot password? Click here to reset