[Feature] Add Res2Net backbone and converted weights. (#465)

* Add Res2Net from mmdet, and change it to mmcls style. * Align structure with official repo * Support `deep_stem` and `avg_down` option * Add Res2Net configs * Add metafile&README and update model zoo * Add unit tests * Imporve docstring. * Improve according to comments.
2025-06-03 21:53:55 +08:00 · 2021-10-20 16:34:22 +08:00 · 2021-10-20 16:34:22 +08:00 · 77a3834531
commit 77a3834531
parent f68f17e9bb
17 changed files with 605 additions and 5 deletions
--- a/configs/_base_/models/res2net101-w26-s4.py
+++ b/configs/_base_/models/res2net101-w26-s4.py
@ -0,0 +1,18 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='Res2Net',
+        depth=101,
+        scales=4,
+        base_width=26,
+        deep_stem=False,
+        avg_down=False,
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5),
+    ))
--- a/configs/_base_/models/res2net50-w14-s8.py
+++ b/configs/_base_/models/res2net50-w14-s8.py
@ -0,0 +1,18 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='Res2Net',
+        depth=50,
+        scales=8,
+        base_width=14,
+        deep_stem=False,
+        avg_down=False,
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5),
+    ))
--- a/configs/_base_/models/res2net50-w26-s4.py
+++ b/configs/_base_/models/res2net50-w26-s4.py
@ -0,0 +1,18 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='Res2Net',
+        depth=50,
+        scales=4,
+        base_width=26,
+        deep_stem=False,
+        avg_down=False,
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5),
+    ))
--- a/configs/_base_/models/res2net50-w26-s6.py
+++ b/configs/_base_/models/res2net50-w26-s6.py
@ -0,0 +1,18 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='Res2Net',
+        depth=50,
+        scales=6,
+        base_width=26,
+        deep_stem=False,
+        avg_down=False,
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5),
+    ))
--- a/configs/_base_/models/res2net50-w26-s8.py
+++ b/configs/_base_/models/res2net50-w26-s8.py
@ -0,0 +1,18 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='Res2Net',
+        depth=50,
+        scales=8,
+        base_width=26,
+        deep_stem=False,
+        avg_down=False,
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5),
+    ))
--- a/configs/_base_/models/res2net50-w48-s2.py
+++ b/configs/_base_/models/res2net50-w48-s2.py
@ -0,0 +1,18 @@
+model = dict(
+    type='ImageClassifier',
+    backbone=dict(
+        type='Res2Net',
+        depth=50,
+        scales=2,
+        base_width=48,
+        deep_stem=False,
+        avg_down=False,
+    ),
+    neck=dict(type='GlobalAveragePooling'),
+    head=dict(
+        type='LinearClsHead',
+        num_classes=1000,
+        in_channels=2048,
+        loss=dict(type='CrossEntropyLoss', loss_weight=1.0),
+        topk=(1, 5),
+    ))
--- a/configs/res2net/README.md
+++ b/configs/res2net/README.md
@ -0,0 +1,30 @@
+# Res2Net: A New Multi-scale Backbone Architecture
+<!-- {Res2Net} -->
+
+## Introduction
+
+<!-- [ALGORITHM] -->
+
+```latex
+@article{gao2019res2net,
+  title={Res2Net: A New Multi-scale Backbone Architecture},
+  author={Gao, Shang-Hua and Cheng, Ming-Ming and Zhao, Kai and Zhang, Xin-Yu and Yang, Ming-Hsuan and Torr, Philip},
+  journal={IEEE TPAMI},
+  year={2021},
+  doi={10.1109/TPAMI.2019.2938758},
+}
+```
+
+## Pretrain model
+
+The pre-trained models are converted from [official repo](https://github.com/Res2Net/Res2Net-PretrainedModels).
+
+### ImageNet 1k
+
+|        Model          | resolution  | Params(M) |  Flops(G) | Top-1 (%) | Top-5 (%) | Download |
+|:---------------------:|:-----------:|:---------:|:---------:|:---------:|:---------:|:--------:|
+|  Res2Net-50-14w-8s\*  |   224x224   |   25.06   |    4.22   |   78.14   |   93.85   | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w14-s8_3rdparty_8xb32_in1k_20210927-bc967bf1.pth)|
+|  Res2Net-50-26w-8s\*  |   224x224   |   48.40   |    8.39   |   79.20   |   94.36   | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w26-s8_3rdparty_8xb32_in1k_20210927-f547a94b.pth)|
+|  Res2Net-101-26w-4s\* |   224x224   |   45.21   |    8.12   |   79.19   |   94.44   | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net101-w26-s4_3rdparty_8xb32_in1k_20210927-870b6c36.pth)|
+
+*Models with \* are converted from other repos.*
--- a/configs/res2net/metafile.yml
+++ b/configs/res2net/metafile.yml
@ -0,0 +1,67 @@
+Collections:
+  - Name: Res2Net
+    Metadata:
+      Training Data: ImageNet-1k
+      Training Techniques:
+        - SGD with Momentum
+        - Weight Decay
+      Architecture:
+        - Batch Normalization
+        - Convolution
+        - Global Average Pooling
+        - ReLU
+        - Res2Net Block
+    Paper:
+      Title: 'Res2Net: A New Multi-scale Backbone Architecture'
+      URL: https://arxiv.org/pdf/1904.01169.pdf
+    README: configs/res2net/README.md
+
+Models:
+  - Name: res2net50-w14-s8_3rdparty_8xb32_in1k
+    Metadata:
+      FLOPs: 4220000000
+      Parameters: 25060000
+    In Collection: Res2Net
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 78.14
+          Top 5 Accuracy: 93.85
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w14-s8_3rdparty_8xb32_in1k_20210927-bc967bf1.pth
+    Converted From:
+      Weights: https://1drv.ms/u/s!AkxDDnOtroRPdOTqhF8ne_aakDI?e=EVb8Ri
+      Code: https://github.com/Res2Net/Res2Net-PretrainedModels/blob/master/res2net.py#L221
+    Config: configs/res2net/res2net50-w14-s8_8xb32_in1k.py
+  - Name: res2net50-w26-s8_3rdparty_8xb32_in1k
+    Metadata:
+      FLOPs: 8390000000
+      Parameters: 48400000
+    In Collection: Res2Net
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 79.20
+          Top 5 Accuracy: 94.36
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w26-s8_3rdparty_8xb32_in1k_20210927-f547a94b.pth
+    Converted From:
+      Weights: https://1drv.ms/u/s!AkxDDnOtroRPdTrAd_Afzc26Z7Q?e=slYqsR
+      Code: https://github.com/Res2Net/Res2Net-PretrainedModels/blob/master/res2net.py#L201
+    Config: configs/res2net/res2net50-w26-s8_8xb32_in1k.py
+  - Name: res2net101-w26-s4_3rdparty_8xb32_in1k
+    Metadata:
+      FLOPs: 8120000000
+      Parameters: 45210000
+    In Collection: Res2Net
+    Results:
+      - Dataset: ImageNet-1k
+        Metrics:
+          Top 1 Accuracy: 79.19
+          Top 5 Accuracy: 94.44
+        Task: Image Classification
+    Weights: https://download.openmmlab.com/mmclassification/v0/res2net/res2net101-w26-s4_3rdparty_8xb32_in1k_20210927-870b6c36.pth
+    Converted From:
+      Weights: https://1drv.ms/u/s!AkxDDnOtroRPcJRgTLkahL0cFYw?e=nwbnic
+      Code: https://github.com/Res2Net/Res2Net-PretrainedModels/blob/master/res2net.py#L181
+    Config: configs/res2net/res2net101-w26-s4_8xb32_in1k.py
--- a/configs/res2net/res2net101-w26-s4_8xb32_in1k.py
+++ b/configs/res2net/res2net101-w26-s4_8xb32_in1k.py
@ -0,0 +1,5 @@
+_base_ = [
+    '../_base_/models/res2net101-w26-s4.py',
+    '../_base_/datasets/imagenet_bs32_pil_resize.py',
+    '../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
+]
--- a/configs/res2net/res2net50-w14-s8_8xb32_in1k.py
+++ b/configs/res2net/res2net50-w14-s8_8xb32_in1k.py
@ -0,0 +1,5 @@
+_base_ = [
+    '../_base_/models/res2net50-w14-s8.py',
+    '../_base_/datasets/imagenet_bs32_pil_resize.py',
+    '../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
+]
--- a/configs/res2net/res2net50-w26-s8_8xb32_in1k.py
+++ b/configs/res2net/res2net50-w26-s8_8xb32_in1k.py
@ -0,0 +1,5 @@
+_base_ = [
+    '../_base_/models/res2net50-w26-s8.py',
+    '../_base_/datasets/imagenet_bs32_pil_resize.py',
+    '../_base_/schedules/imagenet_bs256.py', '../_base_/default_runtime.py'
+]
--- a/docs/model_zoo.md
+++ b/docs/model_zoo.md
@ -32,6 +32,9 @@ The ResNet family models below are trained by standard data augmentations, i.e.,
 | ResNet-50             | 25.56     | 4.12     | 76.55 | 93.15 | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet50_b32x8_imagenet.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_batch256_imagenet_20200708-cfb998bf.pth) &#124; [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_batch256_imagenet_20200708-cfb998bf.log.json) |
 | ResNet-101            | 44.55     | 7.85     | 78.18 | 94.03 | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet101_b32x8_imagenet.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_batch256_imagenet_20200708-753f3608.pth) &#124; [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_batch256_imagenet_20200708-753f3608.log.json) |
 | ResNet-152            | 60.19     | 11.58    | 78.63 | 94.16 | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/resnet/resnet152_b32x8_imagenet.py) | [model](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_batch256_imagenet_20200708-ec25b1f9.pth) &#124; [log](https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_batch256_imagenet_20200708-ec25b1f9.log.json) |
+| Res2Net-50-14w-8s\*   |   25.06   |   4.22   | 78.14 | 93.85 | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/res2net/res2net50-w14-s8_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w14-s8_3rdparty_8xb32_in1k_20210927-bc967bf1.pth)  &#124; [log]()|
+| Res2Net-50-26w-8s\*   |   48.40   |   8.39   | 79.20 | 94.36 | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/res2net/res2net50-w26-s8_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w26-s8_3rdparty_8xb32_in1k_20210927-f547a94b.pth)  &#124; [log]()|
+| Res2Net-101-26w-4s\*  |   45.21   |   8.12   | 79.19 | 94.44 | [config](https://github.com/open-mmlab/mmclassification/blob/master/configs/res2net/res2net101-w26-s4_8xb32_in1k.py) | [model](https://download.openmmlab.com/mmclassification/v0/res2net/res2net101-w26-s4_3rdparty_8xb32_in1k_20210927-870b6c36.pth)  &#124; [log]()|
 | ResNeSt-50\*           | 27.48     | 5.41     | 81.13 | 95.59 |  | [model](https://download.openmmlab.com/mmclassification/v0/resnest/resnest50_imagenet_converted-1ebf0afe.pth) &#124; [log]() |
 | ResNeSt-101\*          | 48.28     | 10.27    | 82.32 | 96.24 |  | [model](https://download.openmmlab.com/mmclassification/v0/resnest/resnest101_imagenet_converted-032caa52.pth) &#124; [log]() |
 | ResNeSt-200\*          | 70.2      | 17.53    | 82.41 | 96.22 |  | [model](https://download.openmmlab.com/mmclassification/v0/resnest/resnest200_imagenet_converted-581a60f2.pth) &#124; [log]() |
--- a/mmcls/models/backbones/init.py
+++ b/mmcls/models/backbones/init.py
@ -5,6 +5,7 @@ from .mobilenet_v2 import MobileNetV2
 from .mobilenet_v3 import MobileNetV3
 from .regnet import RegNet
 from .repvgg import RepVGG
+from .res2net import Res2Net
 from .resnest import ResNeSt
 from .resnet import ResNet, ResNetV1d
 from .resnet_cifar import ResNet_CIFAR
@ -23,5 +24,5 @@ __all__ = [
    'LeNet5', 'AlexNet', 'VGG', 'RegNet', 'ResNet', 'ResNeXt', 'ResNetV1d',
    'ResNeSt', 'ResNet_CIFAR', 'SEResNet', 'SEResNeXt', 'ShuffleNetV1',
    'ShuffleNetV2', 'MobileNetV2', 'MobileNetV3', 'VisionTransformer',
-    'SwinTransformer', 'TNT', 'RepVGG', 'TIMMBackbone'
+    'SwinTransformer', 'TNT', 'TIMMBackbone', 'Res2Net', 'RepVGG'
 ]
--- a/mmcls/models/backbones/res2net.py
+++ b/mmcls/models/backbones/res2net.py
@ -0,0 +1,306 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import math
+
+import torch
+import torch.nn as nn
+import torch.utils.checkpoint as cp
+from mmcv.cnn import build_conv_layer, build_norm_layer
+from mmcv.runner import ModuleList, Sequential
+
+from ..builder import BACKBONES
+from .resnet import Bottleneck as _Bottleneck
+from .resnet import ResNet
+
+
+class Bottle2neck(_Bottleneck):
+    expansion = 4
+
+    def __init__(self,
+                 in_channels,
+                 out_channels,
+                 scales=4,
+                 base_width=26,
+                 base_channels=64,
+                 stage_type='normal',
+                 **kwargs):
+        """Bottle2neck block for Res2Net."""
+        super(Bottle2neck, self).__init__(in_channels, out_channels, **kwargs)
+        assert scales > 1, 'Res2Net degenerates to ResNet when scales = 1.'
+
+        mid_channels = out_channels // self.expansion
+        width = int(math.floor(mid_channels * (base_width / base_channels)))
+
+        self.norm1_name, norm1 = build_norm_layer(
+            self.norm_cfg, width * scales, postfix=1)
+        self.norm3_name, norm3 = build_norm_layer(
+            self.norm_cfg, self.out_channels, postfix=3)
+
+        self.conv1 = build_conv_layer(
+            self.conv_cfg,
+            self.in_channels,
+            width * scales,
+            kernel_size=1,
+            stride=self.conv1_stride,
+            bias=False)
+        self.add_module(self.norm1_name, norm1)
+
+        if stage_type == 'stage':
+            self.pool = nn.AvgPool2d(
+                kernel_size=3, stride=self.conv2_stride, padding=1)
+
+        self.convs = ModuleList()
+        self.bns = ModuleList()
+        for i in range(scales - 1):
+            self.convs.append(
+                build_conv_layer(
+                    self.conv_cfg,
+                    width,
+                    width,
+                    kernel_size=3,
+                    stride=self.conv2_stride,
+                    padding=self.dilation,
+                    dilation=self.dilation,
+                    bias=False))
+            self.bns.append(
+                build_norm_layer(self.norm_cfg, width, postfix=i + 1)[1])
+
+        self.conv3 = build_conv_layer(
+            self.conv_cfg,
+            width * scales,
+            self.out_channels,
+            kernel_size=1,
+            bias=False)
+        self.add_module(self.norm3_name, norm3)
+
+        self.stage_type = stage_type
+        self.scales = scales
+        self.width = width
+        delattr(self, 'conv2')
+        delattr(self, self.norm2_name)
+
+    def forward(self, x):
+        """Forward function."""
+
+        def _inner_forward(x):
+            identity = x
+
+            out = self.conv1(x)
+            out = self.norm1(out)
+            out = self.relu(out)
+
+            spx = torch.split(out, self.width, 1)
+            sp = self.convs[0](spx[0].contiguous())
+            sp = self.relu(self.bns[0](sp))
+            out = sp
+            for i in range(1, self.scales - 1):
+                if self.stage_type == 'stage':
+                    sp = spx[i]
+                else:
+                    sp = sp + spx[i]
+                sp = self.convs[i](sp.contiguous())
+                sp = self.relu(self.bns[i](sp))
+                out = torch.cat((out, sp), 1)
+
+            if self.stage_type == 'normal' and self.scales != 1:
+                out = torch.cat((out, spx[self.scales - 1]), 1)
+            elif self.stage_type == 'stage' and self.scales != 1:
+                out = torch.cat((out, self.pool(spx[self.scales - 1])), 1)
+
+            out = self.conv3(out)
+            out = self.norm3(out)
+
+            if self.downsample is not None:
+                identity = self.downsample(x)
+
+            out += identity
+
+            return out
+
+        if self.with_cp and x.requires_grad:
+            out = cp.checkpoint(_inner_forward, x)
+        else:
+            out = _inner_forward(x)
+
+        out = self.relu(out)
+
+        return out
+
+
+class Res2Layer(Sequential):
+    """Res2Layer to build Res2Net style backbone.
+
+    Args:
+        block (nn.Module): block used to build ResLayer.
+        inplanes (int): inplanes of block.
+        planes (int): planes of block.
+        num_blocks (int): number of blocks.
+        stride (int): stride of the first block. Default: 1
+        avg_down (bool): Use AvgPool instead of stride conv when
+            downsampling in the bottle2neck. Defaults to True.
+        conv_cfg (dict): dictionary to construct and config conv layer.
+            Default: None
+        norm_cfg (dict): dictionary to construct and config norm layer.
+            Default: dict(type='BN')
+        scales (int): Scales used in Res2Net. Default: 4
+        base_width (int): Basic width of each scale. Default: 26
+    """
+
+    def __init__(self,
+                 block,
+                 in_channels,
+                 out_channels,
+                 num_blocks,
+                 stride=1,
+                 avg_down=True,
+                 conv_cfg=None,
+                 norm_cfg=dict(type='BN'),
+                 scales=4,
+                 base_width=26,
+                 **kwargs):
+        self.block = block
+
+        downsample = None
+        if stride != 1 or in_channels != out_channels:
+            if avg_down:
+                downsample = nn.Sequential(
+                    nn.AvgPool2d(
+                        kernel_size=stride,
+                        stride=stride,
+                        ceil_mode=True,
+                        count_include_pad=False),
+                    build_conv_layer(
+                        conv_cfg,
+                        in_channels,
+                        out_channels,
+                        kernel_size=1,
+                        stride=1,
+                        bias=False),
+                    build_norm_layer(norm_cfg, out_channels)[1],
+                )
+            else:
+                downsample = nn.Sequential(
+                    build_conv_layer(
+                        conv_cfg,
+                        in_channels,
+                        out_channels,
+                        kernel_size=1,
+                        stride=stride,
+                        bias=False),
+                    build_norm_layer(norm_cfg, out_channels)[1],
+                )
+
+        layers = []
+        layers.append(
+            block(
+                in_channels=in_channels,
+                out_channels=out_channels,
+                stride=stride,
+                downsample=downsample,
+                conv_cfg=conv_cfg,
+                norm_cfg=norm_cfg,
+                scales=scales,
+                base_width=base_width,
+                stage_type='stage',
+                **kwargs))
+        in_channels = out_channels
+        for _ in range(1, num_blocks):
+            layers.append(
+                block(
+                    in_channels=in_channels,
+                    out_channels=out_channels,
+                    stride=1,
+                    conv_cfg=conv_cfg,
+                    norm_cfg=norm_cfg,
+                    scales=scales,
+                    base_width=base_width,
+                    **kwargs))
+        super(Res2Layer, self).__init__(*layers)
+
+
+@BACKBONES.register_module()
+class Res2Net(ResNet):
+    """Res2Net backbone.
+
+    A PyTorch implement of : `Res2Net: A New Multi-scale Backbone
+    Architecture <https://arxiv.org/pdf/1904.01169.pdf>`_
+
+    Args:
+        depth (int): Depth of Res2Net, choose from {50, 101, 152}.
+        scales (int): Scales used in Res2Net. Defaults to 4.
+        base_width (int): Basic width of each scale. Defaults to 26.
+        in_channels (int): Number of input image channels. Defaults to 3.
+        num_stages (int): Number of Res2Net stages. Defaults to 4.
+        strides (Sequence[int]): Strides of the first block of each stage.
+            Defaults to ``(1, 2, 2, 2)``.
+        dilations (Sequence[int]): Dilation of each stage.
+            Defaults to ``(1, 1, 1, 1)``.
+        out_indices (Sequence[int]): Output from which stages.
+            Defaults to ``(3, )``.
+        style (str): "pytorch" or "caffe". If set to "pytorch", the stride-two
+            layer is the 3x3 conv layer, otherwise the stride-two layer is
+            the first 1x1 conv layer. Defaults to "pytorch".
+        deep_stem (bool): Replace 7x7 conv in input stem with 3 3x3 conv.
+            Defaults to True.
+        avg_down (bool): Use AvgPool instead of stride conv when
+            downsampling in the bottle2neck. Defaults to True.
+        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
+            -1 means not freezing any parameters. Defaults to -1.
+        norm_cfg (dict): Dictionary to construct and config norm layer.
+            Defaults to ``dict(type='BN', requires_grad=True)``.
+        norm_eval (bool): Whether to set norm layers to eval mode, namely,
+            freeze running stats (mean and var). Note: Effect on Batch Norm
+            and its variants only. Defaults to False.
+        with_cp (bool): Use checkpoint or not. Using checkpoint will save some
+            memory while slowing down the training speed. Defaults to False.
+        zero_init_residual (bool): Whether to use zero init for last norm layer
+            in resblocks to let them behave as identity. Defaults to True.
+        init_cfg (dict or list[dict], optional): Initialization config dict.
+            Defaults to None.
+
+    Example:
+        >>> from mmcls.models import Res2Net
+        >>> import torch
+        >>> model = Res2Net(depth=50,
+        ...                 scales=4,
+        ...                 base_width=26,
+        ...                 out_indices=(0, 1, 2, 3))
+        >>> model.eval()
+        >>> inputs = torch.rand(1, 3, 32, 32)
+        >>> level_outputs = model.forward(inputs)
+        >>> for level_out in level_outputs:
+        ...     print(tuple(level_out.shape))
+        (1, 256, 8, 8)
+        (1, 512, 4, 4)
+        (1, 1024, 2, 2)
+        (1, 2048, 1, 1)
+    """
+
+    arch_settings = {
+        50: (Bottle2neck, (3, 4, 6, 3)),
+        101: (Bottle2neck, (3, 4, 23, 3)),
+        152: (Bottle2neck, (3, 8, 36, 3))
+    }
+
+    def __init__(self,
+                 scales=4,
+                 base_width=26,
+                 style='pytorch',
+                 deep_stem=True,
+                 avg_down=True,
+                 init_cfg=None,
+                 **kwargs):
+        self.scales = scales
+        self.base_width = base_width
+        super(Res2Net, self).__init__(
+            style=style,
+            deep_stem=deep_stem,
+            avg_down=avg_down,
+            init_cfg=init_cfg,
+            **kwargs)
+
+    def make_res_layer(self, **kwargs):
+        return Res2Layer(
+            scales=self.scales,
+            base_width=self.base_width,
+            base_channels=self.base_channels,
+            **kwargs)
--- a/mmcls/models/backbones/resnet.py
+++ b/mmcls/models/backbones/resnet.py
@ -396,10 +396,8 @@ class ResNet(BaseBackbone):
            Default: ``(1, 2, 2, 2)``.
        dilations (Sequence[int]): Dilation of each stage.
            Default: ``(1, 1, 1, 1)``.
-        out_indices (Sequence[int]): Output from which stages. If only one
-            stage is specified, a single tensor (feature map) is returned,
-            otherwise multiple stages are specified, a tuple of tensors will
-            be returned. Default: ``(3, )``.
+        out_indices (Sequence[int]): Output from which stages.
+            Default: ``(3, )``.
        style (str): `pytorch` or `caffe`. If set to "pytorch", the stride-two
            layer is the 3x3 conv layer, otherwise the stride-two layer is
            the first 1x1 conv layer.
--- a/model-index.yml
+++ b/model-index.yml
@ -2,6 +2,7 @@ Import:
  - configs/fp16/metafile.yml
  - configs/mobilenet_v2/metafile.yml
  - configs/resnet/metafile.yml
+  - configs/res2net/metafile.yml
  - configs/resnext/metafile.yml
  - configs/seresnet/metafile.yml
  - configs/shufflenet_v1/metafile.yml
--- a/tests/test_models/test_backbones/test_res2net.py
+++ b/tests/test_models/test_backbones/test_res2net.py
@ -0,0 +1,71 @@
+# Copyright (c) OpenMMLab. All rights reserved.
+import pytest
+import torch
+from mmcv.utils.parrots_wrapper import _BatchNorm
+
+from mmcls.models.backbones import Res2Net
+
+
+def check_norm_state(modules, train_state):
+    """Check if norm layer is in correct train state."""
+    for mod in modules:
+        if isinstance(mod, _BatchNorm):
+            if mod.training != train_state:
+                return False
+    return True
+
+
+def test_resnet_cifar():
+    # Only support depth 50, 101 and 152
+    with pytest.raises(KeyError):
+        Res2Net(depth=18)
+
+    # test the feature map size when depth is 50
+    # and deep_stem=True, avg_down=True
+    model = Res2Net(
+        depth=50, out_indices=(0, 1, 2, 3), deep_stem=True, avg_down=True)
+    model.init_weights()
+    model.train()
+
+    imgs = torch.randn(1, 3, 224, 224)
+    feat = model.stem(imgs)
+    assert feat.shape == (1, 64, 112, 112)
+    feat = model(imgs)
+    assert len(feat) == 4
+    assert feat[0].shape == (1, 256, 56, 56)
+    assert feat[1].shape == (1, 512, 28, 28)
+    assert feat[2].shape == (1, 1024, 14, 14)
+    assert feat[3].shape == (1, 2048, 7, 7)
+
+    # test the feature map size when depth is 101
+    # and deep_stem=False, avg_down=False
+    model = Res2Net(
+        depth=101, out_indices=(0, 1, 2, 3), deep_stem=False, avg_down=False)
+    model.init_weights()
+    model.train()
+
+    imgs = torch.randn(1, 3, 224, 224)
+    feat = model.conv1(imgs)
+    assert feat.shape == (1, 64, 112, 112)
+    feat = model(imgs)
+    assert len(feat) == 4
+    assert feat[0].shape == (1, 256, 56, 56)
+    assert feat[1].shape == (1, 512, 28, 28)
+    assert feat[2].shape == (1, 1024, 14, 14)
+    assert feat[3].shape == (1, 2048, 7, 7)
+
+    # Test Res2Net with first stage frozen
+    frozen_stages = 1
+    model = Res2Net(depth=50, frozen_stages=frozen_stages, deep_stem=False)
+    model.init_weights()
+    model.train()
+    assert check_norm_state([model.norm1], False)
+    for param in model.conv1.parameters():
+        assert param.requires_grad is False
+    for i in range(1, frozen_stages + 1):
+        layer = getattr(model, f'layer{i}')
+        for mod in layer.modules():
+            if isinstance(mod, _BatchNorm):
+                assert mod.training is False
+        for param in layer.parameters():
+            assert param.requires_grad is False