visionlearninggroup / dance Goto Github PK

Hey! I was trying to run this code on my machine, but I am not able to create the conda environment due to package errors.
This is the error I am greeted with when I try to create the environment. Could you provide me with a yaml file for this environment or give me some tips to resolve this error.

Thank you,

`

conda create --name dance --file requirements.txt

Collecting package metadata (current_repodata.json): done

Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.

Collecting package metadata (repodata.json): done

Solving environment: failed

PackagesNotFoundError: The following packages are not available from current channels:

simplejson==3.17.2=pypi_0
tensorboard-plugin-wit==1.6.0.post3=pypi_0
smmap==3.0.4=pypi_0
bravado-core==5.17.0=pypi_0
websocket-client==0.57.0=pypi_0
defcon==0.7.2=pypi_0
dill==0.3.2=pypi_0
cxxfilt==0.2.1=py36h831f99a_1
torchvision==0.4.0=py36_cu92
jsonschema==3.2.0=pypi_0
nvidia-apex==0.1=py36h88da601_1
kiwisolver==1.1.0=py36hc9558a2_0
importlib-metadata==1.6.0=py36h9f0ad1d_0
tornado==6.0.3=py36h516909a_0
bleach==3.2.0=pypi_0
tensorflow==2.0.0=pypi_0
_libgcc_mutex==0.1=conda_forge
rsa==4.0=pypi_0
notebook==6.1.4=pypi_0
oauthlib==3.1.0=pypi_0
densratio==0.2.2=pypi_0
scikit-learn==0.20.4=py36_blas_openblashebff5e3_0
pretrainedmodels==0.7.4=pypi_0
typing-extensions==3.7.4.3=pypi_0
dominate==2.5.1=pypi_0
requests-oauthlib==1.3.0=pypi_0
tensorflow-estimator==2.2.0=pypi_0
webencodings==0.5.1=pypi_0
libgcc-ng==9.3.0=h24d8f2e_16
pytorch==1.2.0=py3.6_cuda9.2.148_cudnn7.6.2_0
icu==58.2=hf484d3e_1000
msgpack-python==0.5.6=pypi_0
tabulate==0.8.5=pypi_0
python-dateutil==2.8.0=py_0
terminado==0.8.3=pypi_0
astor==0.8.0=pypi_0
google-pasta==0.1.8=pypi_0
ipykernel==5.3.4=pypi_0
cachetools==3.1.1=pypi_0
nest-asyncio==1.4.0=pypi_0
argon2-cffi==20.1.0=pypi_0
msgpack-numpy==0.4.4.3=pypi_0
glyphslib==2.4.0=pypi_0
nbclient==0.5.0=pypi_0
scipy==1.4.1=pypi_0
hdbscan==0.8.22=py36hd352d35_1
path==15.0.0=pypi_0
openssl==1.1.1g=h516909a_1
kmeans-pytorch==0.3=pypi_0
backcall==0.2.0=pypi_0
astunparse==1.6.3=pypi_0
tensorboardx==1.8=pypi_0
bravado==10.6.2=pypi_0
gitdb==4.0.5=pypi_0
tk==8.6.9=hed695b0_1002
ptyprocess==0.6.0=pypi_0
gitpython==3.1.8=pypi_0
easydict==1.9=pypi_0
monotonic==1.5=pypi_0
pyrsistent==0.17.3=pypi_0
webcolors==1.11.1=pypi_0
libcblas==3.8.0=11_openblas
jinja2==2.11.2=pypi_0
mistune==0.8.4=pypi_0
efficientnet-pytorch==0.7.0=pypi_0
pywsl==0.1.4=pypi_0
tensorpack==0.9.8=pypi_0
traitlets==4.3.3=pypi_0
async-generator==1.10=pypi_0
jsonref==0.2=pypi_0
pyjwt==1.7.1=pypi_0
ipython-genutils==0.2.0=pypi_0
future==0.18.2=pypi_0
jupyter-client==6.1.7=pypi_0
nbconvert==6.0.3=pypi_0
fontconfig==2.13.1=he4413a7_1000
easydl==2.0.7=pypi_0
send2trash==1.5.0=pypi_0
pexpect==4.8.0=pypi_0
swagger-spec-validator==2.7.3=pypi_0
libuuid==2.32.1=h14c3975_1000
click==7.1.2=pypi_0
liblapack==3.8.0=11_openblas
protobuf==3.9.2=pypi_0
xmltodict==0.12.0=pypi_0
strict-rfc3339==0.7=pypi_0
markupsafe==1.1.1=pypi_0
nbformat==5.0.7=pypi_0
pykeops==1.3=pypi_0
urllib3==1.24.3=pypi_0
ipdb==0.13.3=pypi_0
sip==4.18=py36_1
tensorflow-gpu==2.2.0=pypi_0
jupyterlab-pygments==0.1.1=pypi_0
jupyter-core==4.6.3=pypi_0
mutatormath==3.0.1=pypi_0
google-auth==1.7.1=pypi_0
cycler==0.10.0=py_1
msgpack==0.6.2=pypi_0
pyqt==4.11.4=py36_3
certifi==2020.6.20=py36h9f0ad1d_0
entrypoints==0.3=pypi_0
pytest==5.4.1=py36h9f0ad1d_0
sklearn==0.0=pypi_0
wcwidth==0.1.9=pyh9f0ad1d_0
jsonpatch==1.26=pypi_0
grpcio==1.25.0=pypi_0
neptune-client==0.4.120=pypi_0
fontmath==0.6.0=pypi_0
torchfile==0.1.0=pypi_0
prompt-toolkit==3.0.5=pypi_0
python_abi==3.6=1_cp36m
mlconfig==0.1.0=pypi_0
jsonpointer==2.0=pypi_0
path-py==12.5.0=pypi_0
jedi==0.17.2=pypi_0
openblas==0.3.3=h9ac9557_1001
blas==1.1=openblas
defusedxml==0.6.0=pypi_0
google-auth-oauthlib==0.4.1=pypi_0
matplotlib==3.1.0=py36_0
gast==0.2.2=pypi_0
tensorboard==2.2.2=pypi_0
ipython==7.16.1=pypi_0
werkzeug==0.16.0=pypi_0
visdom==0.1.8.9=pypi_0
absl-py==0.8.1=pypi_0
prometheus-client==0.8.0=pypi_0
pathlib2==2.3.5=pypi_0
seaborn==0.11.0=pypi_0
opt-einsum==3.1.0=pypi_0
pyasn1-modules==0.2.7=pypi_0
libxml2==2.9.9=h13577e0_2
neptune-notebooks==0.0.16=pypi_0
pyasn1==0.4.8=pypi_0
libgomp==9.3.0=h24d8f2e_16
pickleshare==0.7.5=pypi_0
libblas==3.8.0=11_openblas
pyzmq==18.1.1=pypi_0
rfc3987==1.3.8=pypi_0
testpath==0.4.4=pypi_0
ca-certificates==2020.6.20=hecda079_0
parso==0.7.1=pypi_0
pygments==2.6.1=pypi_0
faiss==1.5.3=pypi_0
libiconv==1.15=h516909a_1005
psutil==5.6.5=pypi_0
py3nvml==0.2.6=pypi_0
gputil==1.4.0=pypi_0
termcolor==1.1.0=pypi_0
libfaiss==1.6.3=h49bdc20_1_cuda
matplotlib-base==3.1.0=py36h5f35d83_0
qt==4.8.7=2
fonttools==4.12.1=pypi_0
pandocfilters==1.4.2=pypi_0

Current channels:

https://repo.anaconda.com/pkgs/main/linux-64
https://repo.anaconda.com/pkgs/main/noarch
https://repo.anaconda.com/pkgs/r/linux-64
https://repo.anaconda.com/pkgs/r/noarch

To search for alternate channels that may provide the conda package you're

looking for, navigate to
https://anaconda.org

and use the search bar at the top of the page.

`

    

Dear authors,
Thanks for the code release. I was curious about the fact that accuracy of method DANN (in closed set for office 31) reported in the paper is much higher than accuracies reported for it in other papers like CDAN. So I wanted to know what architecture was used for the classifier and discriminator network along with other hyperparams for your DANN implementation.

Thanks

Hi, thanks very much for submitting the good work. I have some simple questions about code in train_dance.py:

1. in 188-189 lines : ### We do not use memory features present in mini-batch
   feat_mat[:, index_t] = -1 / conf.model.temp
   I know the meaning is calculating the similarity between min-batch and min-batch with current features not memory features, but what's the meaning about -1 / conf.model.temp?

2. in 195-196 lines: loss_nc = conf.train.eta * entropy(torch.cat([out_t, feat_mat, feat_mat2], 1))
   I can't understand what's the effect of direct connection of feat_mat and feat_mat2. why not put the feat_mat2 into the proper index position in feat_mat, as we know, the index of feat_t in different iteration is not same.

Thanks very much and hope to get your reply

Thank you for sharing your codes. I have a question.

Regarding Table 4, Office(10/0/11) seems to have a typo.
Which is correct, Office(10/10/11) or Office(10/0/21)?

In the page of 3, in section 3, the authors have said that "In addition, we utilize domain-specific" batch normalization to ..."; However, I do not found that the code is implemented using domain-specific batch normalization. So I am a little curious about that? Can you explain why?

Hello, Thanks for your interesting and inspiring work~

Recently, [1] and [2] proposed a new metric, harmonic mean of Known ACC and Unknown ACC, to better evaluate the capability on identifying known samples and unknown samples.

Will you consider reporting the results with this metric? Or could you release the model trained with the proposed method(DANCE) ?.

[1] Learning to Detect Open Classes for Universal Domain Adaptation
[2] On the Effectiveness of Image Rotation for Open Set Domain Adaptation

Hi, thanks very much for your wonderful work about UniDA.

I have a question about the test Function in eval.py:

• In the line 32-33. You let the prediction of some samples equal to n_share. But I think this may not be correct in some conditions. For example, In the OPDA setting of Office31, the n_share (10)th class is included in the source label set (0 - 19). And this may make the computation of the unkown class accuracy be wrong (line 38).

I don’t know if I misunderstood the code. Thanks very much and hope to get your reply.

After carefully checking the training code, I have a little question about the feat_mat2 in loss_nc.

How does the feat_mat2 help the performance of neighborhood clustering?

I see when computing the loss_nc the entropy was fed with the concatenation of three values which are out_t, feat_mat and feat_mat2. It turns out that the out_t and feat_mat denote the $W$ and $V$ respectively illustrated by the Sec. 3.2. So I guess the feat_mat2 has the same influence with feat_mat. But it only works in the mini batch data, does it? Is there any reasonable explanation about it? Thanks in advance.

In txt file, I find paths of some images from /research/masaito/, what the difference between images in /research/masaito/ and data/

removed