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

67
Seg_All_In_One_MMSeg/configs/dpt/README.md Normal file
View File

@@ -0,0 +1,67 @@
# DPT
> [Vision Transformer for Dense Prediction](https://arxiv.org/abs/2103.13413)
## Introduction
<!-- [ALGORITHM] -->
<a href="https://github.com/isl-org/DPT">Official Repo</a>
<a href="https://github.com/open-mmlab/mmsegmentation/blob/v0.17.0/mmseg/models/decode_heads/dpt_head.py#L215">Code Snippet</a>
## Abstract
<!-- [ABSTRACT] -->
We introduce dense vision transformers, an architecture that leverages vision transformers in place of convolutional networks as a backbone for dense prediction tasks. We assemble tokens from various stages of the vision transformer into image-like representations at various resolutions and progressively combine them into full-resolution predictions using a convolutional decoder. The transformer backbone processes representations at a constant and relatively high resolution and has a global receptive field at every stage. These properties allow the dense vision transformer to provide finer-grained and more globally coherent predictions when compared to fully-convolutional networks. Our experiments show that this architecture yields substantial improvements on dense prediction tasks, especially when a large amount of training data is available. For monocular depth estimation, we observe an improvement of up to 28% in relative performance when compared to a state-of-the-art fully-convolutional network. When applied to semantic segmentation, dense vision transformers set a new state of the art on ADE20K with 49.02% mIoU. We further show that the architecture can be fine-tuned on smaller datasets such as NYUv2, KITTI, and Pascal Context where it also sets the new state of the art. Our models are available at [this https URL](https://github.com/isl-org/DPT).
<!-- [IMAGE] -->
<div align=center>
<img src="https://user-images.githubusercontent.com/24582831/142901057-00aabea5-dab4-43d3-a14a-5f73eb5dd9b9.png" width="80%"/>
</div>
## Usage
To use other repositories' pre-trained models, it is necessary to convert keys.
We provide a script [`vit2mmseg.py`](../../tools/model_converters/vit2mmseg.py) in the tools directory to convert the key of models from [timm](https://github.com/rwightman/pytorch-image-models/blob/master/timm/models/vision_transformer.py) to MMSegmentation style.
```shell
python tools/model_converters/vit2mmseg.py ${PRETRAIN_PATH} ${STORE_PATH}
```
E.g.
```shell
python tools/model_converters/vit2mmseg.py https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-vitjx/jx_vit_base_p16_224-80ecf9dd.pth pretrain/jx_vit_base_p16_224-80ecf9dd.pth
```
This script convert model from `PRETRAIN_PATH` and store the converted model in `STORE_PATH`.
## Results and models
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | Device | mIoU | mIoU(ms+flip) | config | download |
| ------ | -------- | --------- | ------: | -------- | -------------- | ------ | ----: | ------------: | -------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| DPT | ViT-B | 512x512 | 160000 | 8.09 | 10.41 | V100 | 46.97 | 48.34 | [config](https://github.com/open-mmlab/mmsegmentation/blob/main/configs/dpt/dpt_vit-b16_8xb2-160k_ade20k-512x512.py) | [model](https://download.openmmlab.com/mmsegmentation/v0.5/dpt/dpt_vit-b16_512x512_160k_ade20k/dpt_vit-b16_512x512_160k_ade20k-db31cf52.pth) \| [log](https://download.openmmlab.com/mmsegmentation/v0.5/dpt/dpt_vit-b16_512x512_160k_ade20k/dpt_vit-b16_512x512_160k_ade20k-20210809_172025.log.json) |
## Citation
```bibtex
@article{dosoViTskiy2020,
title={An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale},
author={DosoViTskiy, Alexey and Beyer, Lucas and Kolesnikov, Alexander and Weissenborn, Dirk and Zhai, Xiaohua and Unterthiner, Thomas and Dehghani, Mostafa and Minderer, Matthias and Heigold, Georg and Gelly, Sylvain and Uszkoreit, Jakob and Houlsby, Neil},
journal={arXiv preprint arXiv:2010.11929},
year={2020}
}
@article{Ranftl2021,
author = {Ren\'{e} Ranftl and Alexey Bochkovskiy and Vladlen Koltun},
title = {Vision Transformers for Dense Prediction},
journal = {ArXiv preprint},
year = {2021},
}
```

40
Seg_All_In_One_MMSeg/configs/dpt/dpt_vit-b16_8xb2-160k_ade20k-512x512.py Normal file
View File

@@ -0,0 +1,40 @@
_base_ = [
'../_base_/models/dpt_vit-b16.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)
# AdamW optimizer, no weight decay for position embedding & layer norm
# in backbone
optim_wrapper = dict(
_delete_=True,
type='OptimWrapper',
optimizer=dict(
type='AdamW', lr=0.00006, betas=(0.9, 0.999), weight_decay=0.01),
# 自定义参数
paramwise_cfg=dict(
custom_keys={
'pos_embed': dict(decay_mult=0.), # 位置嵌入（positional embeddings）。decay_mult=0. 意味着对这些嵌入不应用权重衰减
'cls_token': dict(decay_mult=0.), # 是在某些模型（如 Transformer 或 BERT）中，用于分类任务的特定 token
'norm': dict(decay_mult=0.) # 对归一化层的参数也禁用了权重衰减
}))
param_scheduler = [
dict(
type='LinearLR', start_factor=1e-6, by_epoch=False, begin=0, end=1500),
dict(
type='PolyLR',
eta_min=0.0,
power=1.0,
begin=1500,
end=160000,
by_epoch=False,
)
]
# By default, models are trained on 8 GPUs with 2 images per GPU
train_dataloader = dict(batch_size=2, num_workers=2)
val_dataloader = dict(batch_size=1, num_workers=4)
test_dataloader = val_dataloader

37
Seg_All_In_One_MMSeg/configs/dpt/metafile.yaml Normal file
View File

@@ -0,0 +1,37 @@
Collections:
- Name: DPT
License: Apache License 2.0
Metadata:
Training Data:
- ADE20K
Paper:
Title: Vision Transformer for Dense Prediction
URL: https://arxiv.org/abs/2103.13413
README: configs/dpt/README.md
Frameworks:
- PyTorch
Models:
- Name: dpt_vit-b16_8xb2-160k_ade20k-512x512
In Collection: DPT
Results:
Task: Semantic Segmentation
Dataset: ADE20K
Metrics:
mIoU: 46.97
mIoU(ms+flip): 48.34
Config: configs/dpt/dpt_vit-b16_8xb2-160k_ade20k-512x512.py
Metadata:
Training Data: ADE20K
Batch Size: 16
Architecture:
- ViT-B
- DPT
Training Resources: 8x V100 GPUS
Memory (GB): 8.09
Weights: https://download.openmmlab.com/mmsegmentation/v0.5/dpt/dpt_vit-b16_512x512_160k_ade20k/dpt_vit-b16_512x512_160k_ade20k-db31cf52.pth
Training log: https://download.openmmlab.com/mmsegmentation/v0.5/dpt/dpt_vit-b16_512x512_160k_ade20k/dpt_vit-b16_512x512_160k_ade20k-20210809_172025.log.json
Paper:
Title: Vision Transformer for Dense Prediction
URL: https://arxiv.org/abs/2103.13413
Code: https://github.com/open-mmlab/mmsegmentation/blob/v0.17.0/mmseg/models/decode_heads/dpt_head.py#L215
Framework: PyTorch

95
Seg_All_In_One_MMSeg/configs/dpt/my_dpt_vit_g1-40k_check_4000_my_dataset_model-1024x1024.py Normal file
View File

@@ -0,0 +1,95 @@
_base_ = [
'../_base_/models/dpt_vit-b16.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 = (1024, 1024)
data_preprocessor = dict(
size=(1024, 1024),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
)
model = dict(
data_preprocessor=dict(
size=(1024, 1024),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
),
pretrained='./My_Local_Model/pretrain/vit-b16_p16_224-80ecf9dd.pth',
decode_head=dict(
loss_decode=dict(
type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0
),
num_classes=36,
),
)
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,
),
paramwise_cfg=dict(
pos_embed=dict(
decay_mult=0.0,
),
cls_token=dict(
decay_mult=0.0,
),
norm=dict(
decay_mult=0.0,
),
),
)
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,
),
]

101
Seg_All_In_One_MMSeg/configs/dpt/my_dpt_vit_g1_b1-40k_check_4000_my_dataset_model-1024x1024.py Normal file
View File

@@ -0,0 +1,101 @@
_base_ = [
'../_base_/models/dpt_vit-b16.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 = (1024, 1024)
data_preprocessor = dict(
size=(1024, 1024),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
)
model = dict(
data_preprocessor=dict(
size=(1024, 1024),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
),
pretrained='./My_Local_Model/pretrain/vit-b16_p16_224-80ecf9dd.pth',
decode_head=dict(
loss_decode=dict(
type='DiceLoss',
use_sigmoid=False,
loss_weight=1.0,
),
num_classes=36,
),
)
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,
),
paramwise_cfg=dict(
pos_embed=dict(
decay_mult=0.0,
),
cls_token=dict(
decay_mult=0.0,
),
norm=dict(
decay_mult=0.0,
),
),
)
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,
),
]
train_dataloader = dict(
batch_size=1,
)

101
Seg_All_In_One_MMSeg/configs/dpt/my_dpt_vit_g1_b1-40k_check_4000_my_dataset_model-512x512.py Normal file
View File

@@ -0,0 +1,101 @@
_base_ = [
'../_base_/models/dpt_vit-b16.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(
data_preprocessor=dict(
size=(512, 512),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
),
pretrained='./My_Local_Model/pretrain/vit-b16_p16_224-80ecf9dd.pth',
decode_head=dict(
loss_decode=dict(
type='DiceLoss',
use_sigmoid=False,
loss_weight=1.0,
),
num_classes=36,
),
)
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,
),
paramwise_cfg=dict(
pos_embed=dict(
decay_mult=0.0,
),
cls_token=dict(
decay_mult=0.0,
),
norm=dict(
decay_mult=0.0,
),
),
)
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,
),
]
train_dataloader = dict(
batch_size=28,
)

101
Seg_All_In_One_MMSeg/configs/dpt/my_dpt_vit_g1_b4-40k_check_4000_my_dataset_model-1024x1024.py Normal file
View File

@@ -0,0 +1,101 @@
_base_ = [
'../_base_/models/dpt_vit-b16.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 = (1024, 1024)
data_preprocessor = dict(
size=(1024, 1024),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
)
model = dict(
data_preprocessor=dict(
size=(1024, 1024),
mean=[
94.94709810464303,
61.72942233949928,
75.93763705236906,
],
std=[
44.005506081132594,
42.69595666984776,
44.99354156225523,
],
bgr_to_rgb=False,
),
pretrained='./My_Local_Model/pretrain/vit-b16_p16_224-80ecf9dd.pth',
decode_head=dict(
loss_decode=dict(
type='DiceLoss',
use_sigmoid=False,
loss_weight=1.0,
),
num_classes=36,
),
)
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,
),
paramwise_cfg=dict(
pos_embed=dict(
decay_mult=0.0,
),
cls_token=dict(
decay_mult=0.0,
),
norm=dict(
decay_mult=0.0,
),
),
)
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,
),
]
train_dataloader = dict(
batch_size=4,
)