python train.py --backbone mobilenet --lr 0.007 --workers 1 --epochs 5 --batch-size 4 --gpu-ids 0 --checkname deeplab-mobilenet
python demo.py --inpath Test --ckpt run/buildings/deeplab-mobilenet-tif/model_best.pth.tar --backbone mobilenet
/root/anaconda3/envs/tf/bin/python /home/pytorch-deeplab/demo.py --in-path /home/pytorch-deeplab/Test --ckpt /home/pytorch-deeplab/run/buildings/deeplab-mobilenet-tif/model_best.pth.tar --backbone mobilenet
# pytorch-deeplab-xception
**Update on 2018/12/06. Provide model trained on VOC and SBD datasets.**
**Update on 2018/11/24. Release newest version code, which fix some previous issues and also add support for new backbones and multi-gpu training. For previous code, please see in `previous` branch**
### TODO
- [x] Support different backbones
- [x] Support VOC, SBD, Cityscapes and COCO datasets
- [x] Multi-GPU training
| Backbone | train/eval os |mIoU in val |Pretrained Model|
| :-------- | :------------: |:---------: |:--------------:|
| ResNet | 16/16 | 78.43% | [google drive](https://drive.google.com/open?id=1NwcwlWqA-0HqAPk3dSNNPipGMF0iS0Zu) |
| MobileNet | 16/16 | 70.81% | [google drive](https://drive.google.com/open?id=1G9mWafUAj09P4KvGSRVzIsV_U5OqFLdt) |
| DRN | 16/16 | 78.87% | [google drive](https://drive.google.com/open?id=131gZN_dKEXO79NknIQazPJ-4UmRrZAfI) |
### Introduction
This is a PyTorch(0.4.1) implementation of [DeepLab-V3-Plus](https://arxiv.org/pdf/1802.02611). It
can use Modified Aligned Xception and ResNet as backbone. Currently, we train DeepLab V3 Plus
using Pascal VOC 2012, SBD and Cityscapes datasets.
![Results](doc/results.png)
### Installation
The code was tested with Anaconda and Python 3.6. After installing the Anaconda environment:
0. Clone the repo:
```Shell
git clone https://github.com/jfzhang95/pytorch-deeplab-xception.git
cd pytorch-deeplab-xception
```
1. Install dependencies:
For PyTorch dependency, see [pytorch.org](https://pytorch.org/) for more details.
For custom dependencies:
```Shell
pip install matplotlib pillow tensorboardX tqdm
```
### Training
Follow steps below to train your model:
0. Configure your dataset path in [mypath.py](https://github.com/jfzhang95/pytorch-deeplab-xception/blob/master/mypath.py).
1. Input arguments: (see full input arguments via python train.py --help):
```Shell
usage: train.py [-h] [--backbone {resnet,xception,drn,mobilenet}]
[--out-stride OUT_STRIDE] [--dataset {pascal,coco,cityscapes}]
[--use-sbd] [--workers N] [--base-size BASE_SIZE]
[--crop-size CROP_SIZE] [--sync-bn SYNC_BN]
[--freeze-bn FREEZE_BN] [--loss-type {ce,focal}] [--epochs N]
[--start_epoch N] [--batch-size N] [--test-batch-size N]
[--use-balanced-weights] [--lr LR]
[--lr-scheduler {poly,step,cos}] [--momentum M]
[--weight-decay M] [--nesterov] [--no-cuda]
[--gpu-ids GPU_IDS] [--seed S] [--resume RESUME]
[--checkname CHECKNAME] [--ft] [--eval-interval EVAL_INTERVAL]
[--no-val]
```
2. To train deeplabv3+ using Pascal VOC dataset and ResNet as backbone:
```Shell
bash train_voc.sh
```
3. To train deeplabv3+ using COCO dataset and ResNet as backbone:
```Shell
bash train_coco.sh
```
### Acknowledgement
[PyTorch-Encoding](https://github.com/zhanghang1989/PyTorch-Encoding)
[Synchronized-BatchNorm-PyTorch](https://github.com/vacancy/Synchronized-BatchNorm-PyTorch)
[drn](https://github.com/fyu/drn)
没有合适的资源?快使用搜索试试~ 我知道了~
WHU-dataset建筑物数据集及模型
共169个文件
pyc:90个
py:44个
txt:9个
5星 · 超过95%的资源 需积分: 27 8 下载量 87 浏览量
2022-11-22
20:42:56
上传
评论 1
收藏 47.07MB RAR 举报
温馨提示
WHU-dataset建筑物数据集及模型
资源推荐
资源详情
资源评论
收起资源包目录
WHU-dataset建筑物数据集及模型 (169个子文件)
events.out.tfevents.1638586336.DL-PC 40B
events.out.tfevents.1638586460.DL-PC 40B
events.out.tfevents.1638585628.DL-PC 40B
events.out.tfevents.1638585393.DL-PC 40B
events.out.tfevents.1638417188.DL-PC 40B
events.out.tfevents.1638623564.DL-PC 0B
events.out.tfevents.1638586767.DL-PC 0B
events.out.tfevents.1638585755.DL-PC 0B
.gitignore 1KB
.gitignore 50B
pytorch-deeplab.iml 495B
events.out.tfevents.1615787648.LAPTOP-H8PSALT1 40B
LICENSE 1KB
README.md 4KB
results.png 509KB
deeplab_xception.py 16KB
new_train.py 15KB
drn.py 14KB
train.py 14KB
batchnorm.py 13KB
xception.py 11KB
deeplab_resnet.py 11KB
resnet.py 6KB
coco.py 6KB
new_demo.py 5KB
mobilenet.py 5KB
cityscapes.py 5KB
buildings.py 5KB
custom_transforms.py 5KB
demo.py 4KB
pascal.py 4KB
comm.py 4KB
sbd.py 4KB
aspp.py 4KB
utils.py 3KB
combine_dbs.py 3KB
replicate.py 3KB
deeplab.py 3KB
__init__.py 3KB
lr_scheduler.py 3KB
saver.py 2KB
decoder.py 2KB
loss.py 2KB
metrics.py 2KB
image-clip.py 1KB
summaries.py 1KB
calculate_weights.py 985B
0-1.py 850B
unittest.py 834B
mypath.py 797B
findcontours.py 769B
findcontours.py 742B
image-augmentation.py 627B
__init__.py 514B
__init__.py 447B
image-add.py 366B
image-tiff-tif.py 322B
__init__.py 0B
__init__.py 0B
batchnorm.cpython-36.pyc 13KB
batchnorm.cpython-37.pyc 13KB
batchnorm.cpython-38.pyc 13KB
drn.cpython-36.pyc 11KB
drn.cpython-37.pyc 11KB
drn.cpython-38.pyc 10KB
xception.cpython-38.pyc 7KB
xception.cpython-37.pyc 7KB
xception.cpython-36.pyc 7KB
cityscapes.cpython-36.pyc 6KB
coco.cpython-36.pyc 5KB
cityscapes.cpython-37.pyc 5KB
custom_transforms.cpython-37.pyc 5KB
custom_transforms.cpython-36.pyc 5KB
cityscapes.cpython-38.pyc 5KB
coco.cpython-37.pyc 5KB
custom_transforms.cpython-38.pyc 5KB
coco.cpython-38.pyc 5KB
resnet.cpython-37.pyc 5KB
resnet.cpython-36.pyc 5KB
resnet.cpython-38.pyc 5KB
comm.cpython-38.pyc 5KB
comm.cpython-37.pyc 5KB
comm.cpython-36.pyc 5KB
buildings.cpython-36.pyc 4KB
pascal.cpython-36.pyc 4KB
buildings.cpython-37.pyc 4KB
pascal.cpython-37.pyc 4KB
buildings.cpython-38.pyc 4KB
pascal.cpython-38.pyc 4KB
mobilenet.cpython-38.pyc 4KB
mobilenet.cpython-37.pyc 4KB
mobilenet.cpython-36.pyc 4KB
sbd.cpython-36.pyc 4KB
sbd.cpython-37.pyc 4KB
sbd.cpython-38.pyc 4KB
replicate.cpython-38.pyc 3KB
replicate.cpython-37.pyc 3KB
utils.cpython-38.pyc 3KB
replicate.cpython-36.pyc 3KB
utils.cpython-37.pyc 3KB
共 169 条
- 1
- 2
资源评论
- yxldr2023-07-27WHU-dataset建筑物数据集及模型的编制方式简洁明了,易于理解和使用。十七张牌你能秒我2023-08-28大家别看了,这些评论都是机器人,当时远程服务器出问题了,通过上传资源来直接下载的数据,下下来你们也不一定会用,不如直接去官方网站下免费的自己处理
- 明儿去打球2023-07-27该文件的建筑物数据集及模型呈现出良好的应用适用性,使研究者们能够更好地分析和处理建筑物相关的问题。
- 我只匆匆而过2023-07-27这份文件的建筑物数据集及模型选取方面考虑周全,能够为建筑物研究领域的进一步发展提供有力支持。
- 我有多作怪2023-07-27WHU-dataset建筑物数据集及模型的内容丰富,能够满足研究者们的多样化需求。
- 永远的122023-07-27这个文件提供了非常详尽的WHU-dataset建筑物数据集及模型,对建筑物研究者来说是非常有用的资源。
十七张牌你能秒我
- 粉丝: 29
- 资源: 2
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功