admin/Seg_Data_Server
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

first commit

Browse Source
This commit is contained in:
admin
2026-05-20 15:05:35 +08:00
commit ac09b26253
2048 changed files with 189478 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

80
Seg_All_In_One_MMSeg/configs/isanet/README.md Normal file
View File

@@ -0,0 +1,80 @@
# ISANet
> [Interlaced Sparse Self-Attention for Semantic Segmentation](https://arxiv.org/abs/1907.12273)
## Introduction
<!-- [ALGORITHM] -->
<a href="https://github.com/openseg-group/openseg.pytorch">Official Repo</a>
<a href="https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58">Code Snippet</a>
## Abstract
<!-- [ABSTRACT] -->
In this paper, we present a so-called interlaced sparse self-attention approach to improve the efficiency of the \\emph{self-attention} mechanism for semantic segmentation. The main idea is that we factorize the dense affinity matrix as the product of two sparse affinity matrices. There are two successive attention modules each estimating a sparse affinity matrix. The first attention module is used to estimate the affinities within a subset of positions that have long spatial interval distances and the second attention module is used to estimate the affinities within a subset of positions that have short spatial interval distances. These two attention modules are designed so that each position is able to receive the information from all the other positions. In contrast to the original self-attention module, our approach decreases the computation and memory complexity substantially especially when processing high-resolution feature maps. We empirically verify the effectiveness of our approach on six challenging semantic segmentation benchmarks.
<!-- [IMAGE] -->
<div align=center>
<img src="https://user-images.githubusercontent.com/24582831/142901868-03d80da4-b9c0-4df9-8509-5f684ba9dadc.png" width="80%"/>
</div>
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | Device | mIoU | mIoU(ms+flip) | config | download |
| ------ | -------- | --------- | ------- | -------: | -------------- | ------ | ----- | ------------: | -----------------------------------------------------------------------------------------------------------------------------: | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| ISANet | R-50-D8 | 512x1024 | 40000 | 5.869 | 2.91 | V100 | 78.49 | 79.44 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r50-d8_4xb2-40k_cityscapes-512x1024.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x1024_40k_cityscapes/isanet_r50-d8_512x1024_40k_cityscapes_20210901_054739-981bd763.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x1024_40k_cityscapes/isanet_r50-d8_512x1024_40k_cityscapes_20210901_054739.log.json) |
| ISANet | R-50-D8 | 512x1024 | 80000 | 5.869 | 2.91 | V100 | 78.68 | 80.25 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r50-d8_4xb2-80k_cityscapes-512x1024.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x1024_80k_cityscapes/isanet_r50-d8_512x1024_80k_cityscapes_20210901_074202-89384497.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x1024_80k_cityscapes/isanet_r50-d8_512x1024_80k_cityscapes_20210901_074202.log.json) |
| ISANet | R-50-D8 | 769x769 | 40000 | 6.759 | 1.54 | V100 | 78.70 | 80.28 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r50-d8_4xb2-40k_cityscapes-769x769.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_769x769_40k_cityscapes/isanet_r50-d8_769x769_40k_cityscapes_20210903_050200-4ae7e65b.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_769x769_40k_cityscapes/isanet_r50-d8_769x769_40k_cityscapes_20210903_050200.log.json) |
| ISANet | R-50-D8 | 769x769 | 80000 | 6.759 | 1.54 | V100 | 79.29 | 80.53 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r50-d8_4xb2-80k_cityscapes-769x769.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_769x769_80k_cityscapes/isanet_r50-d8_769x769_80k_cityscapes_20210903_101126-99b54519.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_769x769_80k_cityscapes/isanet_r50-d8_769x769_80k_cityscapes_20210903_101126.log.json) |
| ISANet | R-101-D8 | 512x1024 | 40000 | 9.425 | 2.35 | V100 | 79.58 | 81.05 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r101-d8_4xb2-40k_cityscapes-512x1024.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x1024_40k_cityscapes/isanet_r101-d8_512x1024_40k_cityscapes_20210901_145553-293e6bd6.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x1024_40k_cityscapes/isanet_r101-d8_512x1024_40k_cityscapes_20210901_145553.log.json) |
| ISANet | R-101-D8 | 512x1024 | 80000 | 9.425 | 2.35 | V100 | 80.32 | 81.58 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r101-d8_4xb2-80k_cityscapes-512x1024.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x1024_80k_cityscapes/isanet_r101-d8_512x1024_80k_cityscapes_20210901_145243-5b99c9b2.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x1024_80k_cityscapes/isanet_r101-d8_512x1024_80k_cityscapes_20210901_145243.log.json) |
| ISANet | R-101-D8 | 769x769 | 40000 | 10.815 | 0.92 | V100 | 79.68 | 80.95 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r101-d8_4xb2-40k_cityscapes-769x769.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_769x769_40k_cityscapes/isanet_r101-d8_769x769_40k_cityscapes_20210903_111320-509e7224.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_769x769_40k_cityscapes/isanet_r101-d8_769x769_40k_cityscapes_20210903_111320.log.json) |
| ISANet | R-101-D8 | 769x769 | 80000 | 10.815 | 0.92 | V100 | 80.61 | 81.59 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r101-d8_4xb2-80k_cityscapes-769x769.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_769x769_80k_cityscapes/isanet_r101-d8_769x769_80k_cityscapes_20210903_111319-24f71dfa.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_769x769_80k_cityscapes/isanet_r101-d8_769x769_80k_cityscapes_20210903_111319.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | Device | mIoU | mIoU(ms+flip) | config | download |
| ------ | -------- | --------- | ------- | -------: | -------------- | ------ | ----- | ------------: | -------------------------------------------------------------------------------------------------------------------------: | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| ISANet | R-50-D8 | 512x512 | 80000 | 9.0 | 22.55 | V100 | 41.12 | 42.35 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r50-d8_4xb4-80k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_80k_ade20k/isanet_r50-d8_512x512_80k_ade20k_20210903_124557-6ed83a0c.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_80k_ade20k/isanet_r50-d8_512x512_80k_ade20k_20210903_124557.log.json) |
| ISANet | R-50-D8 | 512x512 | 160000 | 9.0 | 22.55 | V100 | 42.59 | 43.07 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r50-d8_4xb4-160k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_160k_ade20k/isanet_r50-d8_512x512_160k_ade20k_20210903_104850-f752d0a3.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_160k_ade20k/isanet_r50-d8_512x512_160k_ade20k_20210903_104850.log.json) |
| ISANet | R-101-D8 | 512x512 | 80000 | 12.562 | 10.56 | V100 | 43.51 | 44.38 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r101-d8_4xb4-80k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_80k_ade20k/isanet_r101-d8_512x512_80k_ade20k_20210903_162056-68b235c2.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_80k_ade20k/isanet_r101-d8_512x512_80k_ade20k_20210903_162056.log.json) |
| ISANet | R-101-D8 | 512x512 | 160000 | 12.562 | 10.56 | V100 | 43.80 | 45.4 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r101-d8_4xb4-160k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_160k_ade20k/isanet_r101-d8_512x512_160k_ade20k_20210903_211431-a7879dcd.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_160k_ade20k/isanet_r101-d8_512x512_160k_ade20k_20210903_211431.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | Device | mIoU | mIoU(ms+flip) | config | download |
| ------ | -------- | --------- | ------- | -------: | -------------- | ------ | ----- | ------------: | --------------------------------------------------------------------------------------------------------------------------: | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| ISANet | R-50-D8 | 512x512 | 20000 | 5.9 | 23.08 | V100 | 76.78 | 77.79 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r50-d8_4xb4-20k_voc12aug-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_20k_voc12aug/isanet_r50-d8_512x512_20k_voc12aug_20210901_164838-79d59b80.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_20k_voc12aug/isanet_r50-d8_512x512_20k_voc12aug_20210901_164838.log.json) |
| ISANet | R-50-D8 | 512x512 | 40000 | 5.9 | 23.08 | V100 | 76.20 | 77.22 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r50-d8_4xb4-40k_voc12aug-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_40k_voc12aug/isanet_r50-d8_512x512_40k_voc12aug_20210901_151349-7d08a54e.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_40k_voc12aug/isanet_r50-d8_512x512_40k_voc12aug_20210901_151349.log.json) |
| ISANet | R-101-D8 | 512x512 | 20000 | 9.465 | 7.42 | V100 | 78.46 | 79.16 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r101-d8_4xb4-20k_voc12aug-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_20k_voc12aug/isanet_r101-d8_512x512_20k_voc12aug_20210901_115805-3ccbf355.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_20k_voc12aug/isanet_r101-d8_512x512_20k_voc12aug_20210901_115805.log.json) |
| ISANet | R-101-D8 | 512x512 | 40000 | 9.465 | 7.42 | V100 | 78.12 | 79.04 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/isanet/isanet_r101-d8_4xb4-40k_voc12aug-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_40k_voc12aug/isanet_r101-d8_512x512_40k_voc12aug_20210901_145814-bc71233b.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_40k_voc12aug/isanet_r101-d8_512x512_40k_voc12aug_20210901_145814.log.json) |
## Citation
```bibetex
@article{huang2019isa,
title={Interlaced Sparse Self-Attention for Semantic Segmentation},
author={Huang, Lang and Yuan, Yuhui and Guo, Jianyuan and Zhang, Chao and Chen, Xilin and Wang, Jingdong},
journal={arXiv preprint arXiv:1907.12273},
year={2019}
}
```
The technical report above is also presented at:
```bibetex
@article{yuan2021ocnet,
title={OCNet: Object Context for Semantic Segmentation},
author={Yuan, Yuhui and Huang, Lang and Guo, Jianyuan and Zhang, Chao and Chen, Xilin and Wang, Jingdong},
journal={International Journal of Computer Vision},
pages={1--24},
year={2021},
publisher={Springer}
}
```

2
Seg_All_In_One_MMSeg/configs/isanet/isanet_r101-d8_4xb2-40k_cityscapes-512x1024.py Normal file
View File

@@ -0,0 +1,2 @@
_base_ = './isanet_r50-d8_4xb2-40k_cityscapes-512x1024.py'
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

2
Seg_All_In_One_MMSeg/configs/isanet/isanet_r101-d8_4xb2-40k_cityscapes-769x769.py Normal file
View File

@@ -0,0 +1,2 @@
_base_ = './isanet_r50-d8_4xb2-40k_cityscapes-769x769.py'
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

2
Seg_All_In_One_MMSeg/configs/isanet/isanet_r101-d8_4xb2-80k_cityscapes-512x1024.py Normal file
View File

@@ -0,0 +1,2 @@
_base_ = './isanet_r50-d8_4xb2-80k_cityscapes-512x1024.py'
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

2
Seg_All_In_One_MMSeg/configs/isanet/isanet_r101-d8_4xb2-80k_cityscapes-769x769.py Normal file
View File

@@ -0,0 +1,2 @@
_base_ = './isanet_r50-d8_4xb2-80k_cityscapes-769x769.py'
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

2
Seg_All_In_One_MMSeg/configs/isanet/isanet_r101-d8_4xb4-160k_ade20k-512x512.py Normal file
View File

@@ -0,0 +1,2 @@
_base_ = './isanet_r50-d8_4xb4-160k_ade20k-512x512.py'
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

2
Seg_All_In_One_MMSeg/configs/isanet/isanet_r101-d8_4xb4-20k_voc12aug-512x512.py Normal file
View File

@@ -0,0 +1,2 @@
_base_ = './isanet_r50-d8_4xb4-20k_voc12aug-512x512.py'
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

2
Seg_All_In_One_MMSeg/configs/isanet/isanet_r101-d8_4xb4-40k_voc12aug-512x512.py Normal file
View File

@@ -0,0 +1,2 @@
_base_ = './isanet_r50-d8_4xb4-40k_voc12aug-512x512.py'
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

2
Seg_All_In_One_MMSeg/configs/isanet/isanet_r101-d8_4xb4-80k_ade20k-512x512.py Normal file
View File

@@ -0,0 +1,2 @@
_base_ = './isanet_r50-d8_4xb4-80k_ade20k-512x512.py'
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

7
Seg_All_In_One_MMSeg/configs/isanet/isanet_r50-d8_4xb2-40k_cityscapes-512x1024.py Normal file
View File

@@ -0,0 +1,7 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py', '../_base_/datasets/cityscapes.py',
'../_base_/default_runtime.py', '../_base_/schedules/schedule_40k.py'
]
crop_size = (512, 1024)
data_preprocessor = dict(size=crop_size)
model = dict(data_preprocessor=data_preprocessor)

12
Seg_All_In_One_MMSeg/configs/isanet/isanet_r50-d8_4xb2-40k_cityscapes-769x769.py Normal file
View File

@@ -0,0 +1,12 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py',
'../_base_/datasets/cityscapes_769x769.py', '../_base_/default_runtime.py',
'../_base_/schedules/schedule_40k.py'
]
crop_size = (769, 769)
data_preprocessor = dict(size=crop_size)
model = dict(
data_preprocessor=data_preprocessor,
decode_head=dict(align_corners=True),
auxiliary_head=dict(align_corners=True),
test_cfg=dict(mode='slide', crop_size=(769, 769), stride=(513, 513)))

7
Seg_All_In_One_MMSeg/configs/isanet/isanet_r50-d8_4xb2-80k_cityscapes-512x1024.py Normal file
View File

@@ -0,0 +1,7 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py', '../_base_/datasets/cityscapes.py',
'../_base_/default_runtime.py', '../_base_/schedules/schedule_80k.py'
]
crop_size = (512, 1024)
data_preprocessor = dict(size=crop_size)
model = dict(data_preprocessor=data_preprocessor)

12
Seg_All_In_One_MMSeg/configs/isanet/isanet_r50-d8_4xb2-80k_cityscapes-769x769.py Normal file
View File

@@ -0,0 +1,12 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py',
'../_base_/datasets/cityscapes_769x769.py', '../_base_/default_runtime.py',
'../_base_/schedules/schedule_80k.py'
]
crop_size = (769, 769)
data_preprocessor = dict(size=crop_size)
model = dict(
data_preprocessor=data_preprocessor,
decode_head=dict(align_corners=True),
auxiliary_head=dict(align_corners=True),
test_cfg=dict(mode='slide', crop_size=(769, 769), stride=(513, 513)))

10
Seg_All_In_One_MMSeg/configs/isanet/isanet_r50-d8_4xb4-160k_ade20k-512x512.py Normal file
View File

@@ -0,0 +1,10 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py', '../_base_/datasets/ade20k.py',
'../_base_/default_runtime.py', '../_base_/schedules/schedule_160k.py'
]
crop_size = (512, 512)
data_preprocessor = dict(size=crop_size)
model = dict(
data_preprocessor=data_preprocessor,
decode_head=dict(num_classes=150),
auxiliary_head=dict(num_classes=150))

11
Seg_All_In_One_MMSeg/configs/isanet/isanet_r50-d8_4xb4-20k_voc12aug-512x512.py Normal file
View File

@@ -0,0 +1,11 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py',
'../_base_/datasets/pascal_voc12_aug.py', '../_base_/default_runtime.py',
'../_base_/schedules/schedule_20k.py'
]
crop_size = (512, 512)
data_preprocessor = dict(size=crop_size)
model = dict(
data_preprocessor=data_preprocessor,
decode_head=dict(num_classes=21),
auxiliary_head=dict(num_classes=21))

11
Seg_All_In_One_MMSeg/configs/isanet/isanet_r50-d8_4xb4-40k_voc12aug-512x512.py Normal file
View File

@@ -0,0 +1,11 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py',
'../_base_/datasets/pascal_voc12_aug.py', '../_base_/default_runtime.py',
'../_base_/schedules/schedule_40k.py'
]
crop_size = (512, 512)
data_preprocessor = dict(size=crop_size)
model = dict(
data_preprocessor=data_preprocessor,
decode_head=dict(num_classes=21),
auxiliary_head=dict(num_classes=21))

10
Seg_All_In_One_MMSeg/configs/isanet/isanet_r50-d8_4xb4-80k_ade20k-512x512.py Normal file
View File

@@ -0,0 +1,10 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py', '../_base_/datasets/ade20k.py',
'../_base_/default_runtime.py', '../_base_/schedules/schedule_80k.py'
]
crop_size = (512, 512)
data_preprocessor = dict(size=crop_size)
model = dict(
data_preprocessor=data_preprocessor,
decode_head=dict(num_classes=150),
auxiliary_head=dict(num_classes=150))

399
Seg_All_In_One_MMSeg/configs/isanet/metafile.yaml Normal file
View File

@@ -0,0 +1,399 @@
Collections:
- Name: ISANet
License: Apache License 2.0
Metadata:
Training Data:
- Cityscapes
- ADE20K
- Pascal VOC 2012 + Aug
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
README: configs/isanet/README.md
Frameworks:
- PyTorch
Models:
- Name: isanet_r50-d8_4xb2-40k_cityscapes-512x1024
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Cityscapes
Metrics:
mIoU: 78.49
mIoU(ms+flip): 79.44
Config: configs/isanet/isanet_r50-d8_4xb2-40k_cityscapes-512x1024.py
Metadata:
Training Data: Cityscapes
Batch Size: 8
Architecture:
- R-50-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 5.869
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x1024_40k_cityscapes/isanet_r50-d8_512x1024_40k_cityscapes_20210901_054739-981bd763.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x1024_40k_cityscapes/isanet_r50-d8_512x1024_40k_cityscapes_20210901_054739.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r50-d8_4xb2-80k_cityscapes-512x1024
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Cityscapes
Metrics:
mIoU: 78.68
mIoU(ms+flip): 80.25
Config: configs/isanet/isanet_r50-d8_4xb2-80k_cityscapes-512x1024.py
Metadata:
Training Data: Cityscapes
Batch Size: 8
Architecture:
- R-50-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 5.869
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x1024_80k_cityscapes/isanet_r50-d8_512x1024_80k_cityscapes_20210901_074202-89384497.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x1024_80k_cityscapes/isanet_r50-d8_512x1024_80k_cityscapes_20210901_074202.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r50-d8_4xb2-40k_cityscapes-769x769
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Cityscapes
Metrics:
mIoU: 78.7
mIoU(ms+flip): 80.28
Config: configs/isanet/isanet_r50-d8_4xb2-40k_cityscapes-769x769.py
Metadata:
Training Data: Cityscapes
Batch Size: 8
Architecture:
- R-50-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 6.759
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_769x769_40k_cityscapes/isanet_r50-d8_769x769_40k_cityscapes_20210903_050200-4ae7e65b.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_769x769_40k_cityscapes/isanet_r50-d8_769x769_40k_cityscapes_20210903_050200.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r50-d8_4xb2-80k_cityscapes-769x769
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Cityscapes
Metrics:
mIoU: 79.29
mIoU(ms+flip): 80.53
Config: configs/isanet/isanet_r50-d8_4xb2-80k_cityscapes-769x769.py
Metadata:
Training Data: Cityscapes
Batch Size: 8
Architecture:
- R-50-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 6.759
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_769x769_80k_cityscapes/isanet_r50-d8_769x769_80k_cityscapes_20210903_101126-99b54519.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_769x769_80k_cityscapes/isanet_r50-d8_769x769_80k_cityscapes_20210903_101126.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r101-d8_4xb2-40k_cityscapes-512x1024
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Cityscapes
Metrics:
mIoU: 79.58
mIoU(ms+flip): 81.05
Config: configs/isanet/isanet_r101-d8_4xb2-40k_cityscapes-512x1024.py
Metadata:
Training Data: Cityscapes
Batch Size: 8
Architecture:
- R-101-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 9.425
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x1024_40k_cityscapes/isanet_r101-d8_512x1024_40k_cityscapes_20210901_145553-293e6bd6.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x1024_40k_cityscapes/isanet_r101-d8_512x1024_40k_cityscapes_20210901_145553.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r101-d8_4xb2-80k_cityscapes-512x1024
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Cityscapes
Metrics:
mIoU: 80.32
mIoU(ms+flip): 81.58
Config: configs/isanet/isanet_r101-d8_4xb2-80k_cityscapes-512x1024.py
Metadata:
Training Data: Cityscapes
Batch Size: 8
Architecture:
- R-101-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 9.425
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x1024_80k_cityscapes/isanet_r101-d8_512x1024_80k_cityscapes_20210901_145243-5b99c9b2.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x1024_80k_cityscapes/isanet_r101-d8_512x1024_80k_cityscapes_20210901_145243.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r101-d8_4xb2-40k_cityscapes-769x769
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Cityscapes
Metrics:
mIoU: 79.68
mIoU(ms+flip): 80.95
Config: configs/isanet/isanet_r101-d8_4xb2-40k_cityscapes-769x769.py
Metadata:
Training Data: Cityscapes
Batch Size: 8
Architecture:
- R-101-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 10.815
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_769x769_40k_cityscapes/isanet_r101-d8_769x769_40k_cityscapes_20210903_111320-509e7224.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_769x769_40k_cityscapes/isanet_r101-d8_769x769_40k_cityscapes_20210903_111320.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r101-d8_4xb2-80k_cityscapes-769x769
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Cityscapes
Metrics:
mIoU: 80.61
mIoU(ms+flip): 81.59
Config: configs/isanet/isanet_r101-d8_4xb2-80k_cityscapes-769x769.py
Metadata:
Training Data: Cityscapes
Batch Size: 8
Architecture:
- R-101-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 10.815
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_769x769_80k_cityscapes/isanet_r101-d8_769x769_80k_cityscapes_20210903_111319-24f71dfa.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_769x769_80k_cityscapes/isanet_r101-d8_769x769_80k_cityscapes_20210903_111319.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r50-d8_4xb4-80k_ade20k-512x512
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 41.12
mIoU(ms+flip): 42.35
Config: configs/isanet/isanet_r50-d8_4xb4-80k_ade20k-512x512.py
Metadata:
Training Data: ADE20K
Batch Size: 16
Architecture:
- R-50-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 9.0
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_80k_ade20k/isanet_r50-d8_512x512_80k_ade20k_20210903_124557-6ed83a0c.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_80k_ade20k/isanet_r50-d8_512x512_80k_ade20k_20210903_124557.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r50-d8_4xb4-160k_ade20k-512x512
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 42.59
mIoU(ms+flip): 43.07
Config: configs/isanet/isanet_r50-d8_4xb4-160k_ade20k-512x512.py
Metadata:
Training Data: ADE20K
Batch Size: 16
Architecture:
- R-50-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 9.0
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_160k_ade20k/isanet_r50-d8_512x512_160k_ade20k_20210903_104850-f752d0a3.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_160k_ade20k/isanet_r50-d8_512x512_160k_ade20k_20210903_104850.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r101-d8_4xb4-80k_ade20k-512x512
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 43.51
mIoU(ms+flip): 44.38
Config: configs/isanet/isanet_r101-d8_4xb4-80k_ade20k-512x512.py
Metadata:
Training Data: ADE20K
Batch Size: 16
Architecture:
- R-101-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 12.562
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_80k_ade20k/isanet_r101-d8_512x512_80k_ade20k_20210903_162056-68b235c2.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_80k_ade20k/isanet_r101-d8_512x512_80k_ade20k_20210903_162056.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r101-d8_4xb4-160k_ade20k-512x512
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 43.8
mIoU(ms+flip): 45.4
Config: configs/isanet/isanet_r101-d8_4xb4-160k_ade20k-512x512.py
Metadata:
Training Data: ADE20K
Batch Size: 16
Architecture:
- R-101-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 12.562
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_160k_ade20k/isanet_r101-d8_512x512_160k_ade20k_20210903_211431-a7879dcd.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_160k_ade20k/isanet_r101-d8_512x512_160k_ade20k_20210903_211431.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r50-d8_4xb4-20k_voc12aug-512x512
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Pascal VOC 2012 + Aug
Metrics:
mIoU: 76.78
mIoU(ms+flip): 77.79
Config: configs/isanet/isanet_r50-d8_4xb4-20k_voc12aug-512x512.py
Metadata:
Training Data: Pascal VOC 2012 + Aug
Batch Size: 16
Architecture:
- R-50-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 5.9
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_20k_voc12aug/isanet_r50-d8_512x512_20k_voc12aug_20210901_164838-79d59b80.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_20k_voc12aug/isanet_r50-d8_512x512_20k_voc12aug_20210901_164838.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r50-d8_4xb4-40k_voc12aug-512x512
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Pascal VOC 2012 + Aug
Metrics:
mIoU: 76.2
mIoU(ms+flip): 77.22
Config: configs/isanet/isanet_r50-d8_4xb4-40k_voc12aug-512x512.py
Metadata:
Training Data: Pascal VOC 2012 + Aug
Batch Size: 16
Architecture:
- R-50-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 5.9
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_40k_voc12aug/isanet_r50-d8_512x512_40k_voc12aug_20210901_151349-7d08a54e.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r50-d8_512x512_40k_voc12aug/isanet_r50-d8_512x512_40k_voc12aug_20210901_151349.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r101-d8_4xb4-20k_voc12aug-512x512
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Pascal VOC 2012 + Aug
Metrics:
mIoU: 78.46
mIoU(ms+flip): 79.16
Config: configs/isanet/isanet_r101-d8_4xb4-20k_voc12aug-512x512.py
Metadata:
Training Data: Pascal VOC 2012 + Aug
Batch Size: 16
Architecture:
- R-101-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 9.465
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_20k_voc12aug/isanet_r101-d8_512x512_20k_voc12aug_20210901_115805-3ccbf355.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_20k_voc12aug/isanet_r101-d8_512x512_20k_voc12aug_20210901_115805.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch
- Name: isanet_r101-d8_4xb4-40k_voc12aug-512x512
In Collection: ISANet
Results:
Task: Semantic Segmentation
Dataset: Pascal VOC 2012 + Aug
Metrics:
mIoU: 78.12
mIoU(ms+flip): 79.04
Config: configs/isanet/isanet_r101-d8_4xb4-40k_voc12aug-512x512.py
Metadata:
Training Data: Pascal VOC 2012 + Aug
Batch Size: 16
Architecture:
- R-101-D8
- ISANet
Training Resources: 4x V100 GPUS
Memory (GB): 9.465
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_40k_voc12aug/isanet_r101-d8_512x512_40k_voc12aug_20210901_145814-bc71233b.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/isanet/isanet_r101-d8_512x512_40k_voc12aug/isanet_r101-d8_512x512_40k_voc12aug_20210901_145814.log.json
Paper:
Title: Interlaced Sparse Self-Attention for Semantic Segmentation
URL: https://arxiv.org/abs/1907.12273
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.18.0/mmseg/models/decode_heads/isa_head.py#L58
Framework: PyTorch

105
Seg_All_In_One_MMSeg/configs/isanet/my_isanet_r50_r101_Pre_g1-40k_check_4000_my_dataset_model-512x512-testslide.py Normal file
View File

@@ -0,0 +1,105 @@
_base_ = [
'../_base_/models/isanet_r50-d8.py',
'../_base_/datasets/my_dataset_model.py',
'../_base_/default_runtime.py',
'../_base_/schedules/schedule_40k_check_4000.py',
]
norm_cfg = dict(
type='BN',
)
crop_size = (512, 512)
data_preprocessor = dict(
size=(512, 512),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
)
model = dict(
pretrained='./My_Local_Model/open_mmlab/resnet101_v1c.pth',
backbone=dict(
depth=101,
),
data_preprocessor=dict(
size=(512, 512),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
),
decode_head=dict(
num_classes=36,
loss_decode=dict(
type='DiceLoss',
use_sigmoid=False,
loss_weight=1.0,
),
align_corners=True,
),
auxiliary_head=dict(
num_classes=36,
loss_decode=dict(
type='DiceLoss',
use_sigmoid=False,
loss_weight=0.4,
),
align_corners=True,
),
)
test_cfg = dict(
mode='slide',
crop_size=(512, 512),
stride=(341, 341),
)
optim_wrapper = dict(
type='OptimWrapper',
_delete_=True,
optimizer=dict(
type='AdamW',
lr=0.0001,
weight_decay=0.0005,
),
clip_grad=dict(
max_norm=1,
norm_type=2,
),
)
param_scheduler = [
dict(
type='LinearLR',
start_factor=1e-06,
by_epoch=False,
begin=0,
end=1500,
),
dict(
type='PolyLR',
power=0.9,
begin=1500,
end=40000,
eta_min=1e-05,
by_epoch=False,
),
]