Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly
Sign up
Appearance settings

Commit b9f489e

Browse files
update
1 parent aeb744f commit b9f489e

File tree

1 file changed

+31
-36
lines changed

1 file changed

+31
-36
lines changed

‎readme.md‎

Lines changed: 31 additions & 36 deletions
Original file line numberDiff line numberDiff line change
@@ -20,94 +20,89 @@ Code for our ICML-2020 paper [**Do We Really Need to Access the Source Data? Sou
2020
1. ##### Unsupervised Closed-set Domain Adaptation (UDA) on the Digits dataset
2121
- MNIST -> USPS (**m2u**) SHOT (**cls_par = 0.1**) and SHOT-IM (**cls_par = 0.0**)
2222
```python
23-
cd digit/
24-
python uda_digit.py --dset m2u --gpu_id 0 --output seed2020 --seed 2020 --cls_par 0.0
25-
python uda_digit.py --dset m2u --gpu_id 0 --output seed2020 --seed 2020 --cls_par 0.1
23+
cd digit/
24+
python uda_digit.py --dset m2u --gpu_id 0 --output ckps_digits --cls_par 0.0
25+
python uda_digit.py --dset m2u --gpu_id 0 --output ckps_digits --cls_par 0.1
2626
```
2727

2828
2. ##### Unsupervised Closed-set Domain Adaptation (UDA) on the Office/ Office-Home dataset
2929
- Train model on the source domain **A** (**s = 0**)
3030
```python
3131
cd object/
32-
python image_source.py --trte val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office --max_epoch 30 --s 0
32+
python image_source.py --trte val --da uda --output ckps/source/ --gpu_id 0 --dset office --max_epoch 100 --s 0
3333
```
3434

3535
- Adaptation to other target domains **D and W**, respectively
3636
```python
37-
python image_target.py --savename par0.0 --cls_par 0.0 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office --max_epoch 30 --s 0
38-
python image_target.py --savename par0.3 --cls_par 0.3 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office --max_epoch 30 --s 0
37+
python image_target.py --cls_par 0.3 --da uda --output_src ckps/source/ --output ckps/target/ --gpu_id 0 --dset office --s 0
3938
```
40-
39+
4140
3. ##### Unsupervised Closed-set Domain Adaptation (UDA) on the VisDA-C dataset
4241
- Synthetic-to-real
4342
```python
4443
cd object/
45-
python uda_visda.py --savename par0.0 --cls_par 0.0 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --max_epoch 3
46-
python uda_visda.py --savename par0.3 --cls_par 0.3 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --max_epoch 3
44+
python image_source.py --trte val --output ckps/source/ --da uda --gpu_id 0 --dset VISDA-C --net resnet101 --lr 1e-3 --max_epoch 10 --s 0
45+
python image_target.py --cls_par 0.3 --da uda --dset VISDA-C --gpu_id 0 --s 0 --output_src ckps/source/ --output ckps/target/ --net resnet101 --lr 1e-3
4746
```
4847

4948
4. ##### Unsupervised Partial-set Domain Adaptation (PDA) on the Office-Home dataset
5049
- Train model on the source domain **A** (**s = 0**)
5150
```python
52-
cd object/
53-
python image_source.py --trte val --da pda --output seed2020 --seed 2020 --gpu_id 0 --dset office-home --max_epoch 30 --s 0
51+
cd object/
52+
python image_source.py --trte val --da pda --output ckps/source/ --gpu_id 0 --dset office-home --max_epoch 50 --s 0
5453
```
5554

5655
- Adaptation to other target domains **C and P and R**, respectively
5756
```python
58-
python image_target.py --savename par0.0 --cls_par 0.0 --zz val --da pda --gent '' --threshold 10 --output seed2020 --seed 2020 --gpu_id 0 --dset office-home --max_epoch 30 --s 0
59-
python image_target.py --savename par0.3 --cls_par 0.3 --zz val --da pda --gent '' --threshold 10 --output seed2020 --seed 2020 --gpu_id 0 --dset office-home --max_epoch 30 --s 0
60-
```
61-
57+
python image_target.py --cls_par 0.3 --threshold 10 --da pda --dset office-home --gpu_id 0 --s 0 --output_src ckps/source/ --output ckps/target/
58+
```
59+
6260
5. ##### Unsupervised Open-set Domain Adaptation (ODA) on the Office-Home dataset
6361
- Train model on the source domain **A** (**s = 0**)
6462
```python
65-
cd object/
66-
python image_source.py --trte val --da oda --output seed2020 --seed 2020 --gpu_id 0 --dset office-home --max_epoch 30 --s 0
63+
cd object/
64+
python image_source.py --trte val --da oda --output ckps/source/ --gpu_id 0 --dset office-home --max_epoch 50 --s 0
6765
```
6866

6967
- Adaptation to other target domains **C and P and R**, respectively
7068
```python
71-
python image_target_oda.py --savename par0.0 --cls_par 0.0 --zz val --da oda --output seed2020 --seed 2020 --gpu_id 0 --dset office-home --max_epoch 30 --s 0
72-
python image_target_oda.py --savename par0.3 --cls_par 0.3 --zz val --da oda --output seed2020 --seed 2020 --gpu_id 0 --dset office-home --max_epoch 30 --s 0
69+
python image_target_oda.py --cls_par 0.3 --da oda --dset office-home --gpu_id 0 --s 0 --output_src ckps/source/ --output ckps/target/
7370
```
74-
71+
7572
6. ##### Unsupervised Multi-source Domain Adaptation (MSDA) on the Office-Caltech dataset
7673
- Train model on the source domains **A** (**s = 0**), **C** (**s = 1**), **D** (**s = 2**), respectively
7774
```python
78-
cd object/
79-
python image_source.py --trte val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office-caltech --net resnet101 --max_epoch 30 --s 0
80-
python image_source.py --trte val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office-caltech --net resnet101 --max_epoch 30 --s 1
81-
python image_source.py --trte val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office-caltech --net resnet101 --max_epoch 30 --s 2
75+
cd object/
76+
python image_source.py --trte val --da uda --output ckps/source/ --gpu_id 0 --dset office-caltech --max_epoch 100 --s 0
77+
python image_source.py --trte val --da uda --output ckps/source/ --gpu_id 0 --dset office-caltech --max_epoch 100 --s 1
78+
python image_source.py --trte val --da uda --output ckps/source/ --gpu_id 0 --dset office-caltech --max_epoch 100 --s 2
8279
```
8380

8481
- Adaptation to the target domain **W** (**t = 3**)
8582
```python
86-
python image_target.py --savename par0.3 --cls_par 0.3 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --issave 1 --dset office-caltech --net resnet101 --max_epoch 30 --s 0
87-
python image_target.py --savename par0.3 --cls_par 0.3 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --issave 1 --dset office-caltech --net resnet101 --max_epoch 30 --s 1
88-
python image_target.py --savename par0.3 --cls_par 0.3 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --issave 1 --dset office-caltech --net resnet101 --max_epoch 30 --s 2
89-
python image_multisource.py --savename par0.0 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office-caltech --net resnet101 --max_epoch 30 --t 3
83+
python image_target.py --cls_par 0.3 --da uda --output_src ckps/source/ --output ckps/target/ --gpu_id 0 --dset office --s 0
84+
python image_target.py --cls_par 0.3 --da uda --output_src ckps/source/ --output ckps/target/ --gpu_id 0 --dset office --s 1
85+
python image_target.py --cls_par 0.3 --da uda --output_src ckps/source/ --output ckps/target/ --gpu_id 0 --dset office --s 0
86+
python image_multisource.py --cls_par 0.0 --da uda --dset office-caltech --gpu_id 0 --t 3 --output_src ckps/source/ --output ckps/target/
9087
```
9188

9289
7. ##### Unsupervised Multi-target Domain Adaptation (MTDA) on the Office-Caltech dataset
9390
- Train model on the source domain **A** (**s = 0**)
9491
```python
95-
cd object/
96-
python image_source.py --trte val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office-caltech --net resnet101 --max_epoch 30 --s 0
92+
cd object/
93+
python image_source.py --trte val --da uda --output ckps/source/ --gpu_id 0 --dset office-caltech --max_epoch 100 --s 0
9794
```
9895

9996
- Adaptation to multiple target domains **C and P and R** at the same time
10097
```python
101-
python image_multitarget.py --savename par0.0 --cls_par 0.0 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office-caltech --net resnet101 --max_epoch 30 --s 0
102-
python image_multitarget.py --savename par0.3 --cls_par 0.3 --zz val --da uda --output seed2020 --seed 2020 --gpu_id 0 --dset office-caltech --net resnet101 --max_epoch 30 --s 0
98+
python image_multitarget.py --cls_par 0.3 --da uda --dset office-caltech --gpu_id 0 --s 0 --output_src ckps/source/ --output ckps/target/
10399
```
104-
100+
105101
8. ##### Unsupervised Partial Domain Adaptation (PDA) on the ImageNet-Caltech dataset without source training by ourselves (using the downloaded Pytorch ResNet50 model directly)
106102
- ImageNet -> Caltech (84 classes) [following the protocol in [PADA](https://github.com/thuml/PADA/tree/master/pytorch/data/imagenet-caltech)]
107103
```python
108-
cd object/
109-
python image_pretrained.py --savename par0.0 --cls_par 0.0 --output seed2020 --seed 2020 --gpu_id 0 --max_epoch 30
110-
python image_pretrained.py --savename par0.3 --cls_par 0.3 --output seed2020 --seed 2020 --gpu_id 0 --max_epoch 30
104+
cd object/
105+
python image_pretrained.py --gpu_id 0 --output ckps/target/--cls_par 0.3
111106
```
112107

113108
**Please refer *run.sh*** for all the settings for different methods and scenarios.

0 commit comments

Comments
(0)

AltStyle によって変換されたページ (->オリジナル) /