Res2Net

Introduced by Gao et al. in Res2Net: A New Multi-scale Backbone Architecture

Res2Net is an image model that employs a variation on bottleneck residual blocks. The motivation is to be able to represent features at multiple scales. This is achieved through a novel building block for CNNs that constructs hierarchical residual-like connections within one single residual block. This represents multi-scale features at a granular level and increases the range of receptive fields for each network layer.

Source: Res2Net: A New Multi-scale Backbone Architecture

Read Paper See Code

Papers

Paper	Code	Results	Date	Stars

Tasks

Task	Papers	Share
Speaker Verification	9	21.43%
Semantic Segmentation	2	4.76%
Speaker Recognition	2	4.76%
Computational Efficiency	1	2.38%
Deep Learning	1	2.38%
Speech Enhancement	1	2.38%
Retinal Vessel Segmentation	1	2.38%
Graph Attention	1	2.38%
Classification	1	2.38%

Usage Over Time

This feature is experimental; we are continuously improving our matching algorithm.

Components

Component	Type	Add Remove
Batch Normalization	Normalization
Convolution	Convolutions
Global Average Pooling	Pooling Operations
Kaiming Initialization	Initialization
ReLU	Activation Functions
Res2Net Block	Skip Connection Blocks

Categories

Add Remove

Image Models