rfelixmg / frwgan-eccv18 Goto Github PK
View Code? Open in Web Editor NEWCode for model presented on our paper accepted on European Conference on Computer Vision 2018.
Code for model presented on our paper accepted on European Conference on Computer Vision 2018.
Hi, thanks for the hard work on releasing the code! I'm wondering if there is the code for training the cycle-clswgan. I didn't seem to find it in the repo. Thanks!
Hi,
Can I ask why H can only reach 50%~51.5% instead of 53%?, i strictly run your code without making changes to the parameters.
Best Wishes,
Hi @rfelixmg ,
Thank you for sharing your valuable code.
I run src/sota/awa1/train_gan.sh and I am getting error while saving checkpoints. Could you please guide me in solving this. I am using Google Colab
I will appreciate.
Hi thanks for your work! I have downloaded the Semantic space of CUB ,Can you please provide the Semantic space of FLO dataset? Thank you.
Hi Rafa,
First, thank you again for making your code available to reproduce the experiments.
I managed to closely get the results on CUB, but I couldn't reproduce the results on FLO or AWA.
On FLO, the problem is that I got really bad results, while on AWA one of the scripts throws an error. For reference, on CUB I got H=52.6
In more detail: After setting up the environment, according to the instructions in the README, for each dataset I execute the following:
source src/sota/<dataset_name>/train_gan.sh
mv ./experiments/sota/<dataset_name>/cycle_wgan/0000_TEST_<TIME>/ ./experiments/sota/<dataset_name>/cycle_wgan/gan-model
source src/sota/<dataset_name>/generate_dataset.sh # and choosing the last epoch to generate feature
source src/sota/<dataset_name>/benchmark.sh
mv ./experiments/sota/<dataset_name>/cycle_wgan/0000_classifier_<TIME>/ ./experiments/sota/<dataset_name>/cycle_wgan/classifier
source src/sota/<dataset_name>/tester.sh
For FLO I get the following results: (y(U) =59.7, y(S) =0, H=0)
For AWA I get the following error when running tester.sh
:
FileNotFoundError: [Errno 2] No such file or directory: './experiments/sota/awa1/cycle_wgan/classifier/checkpoint/epoch_80_80/architecture.json'
It looks like some kind of configuration error on the scripts.
Could you kindly assist with this issue?
Many Thanks,
Yuval
Hi Rafa,
Would it be possible to share the code for reproducing the paper results on ImageNet dataset?
Thank you and thanks for all your help so far.
Best,
Yuval
Hi,
Can I ask where can download the architecture file?
Best Wishes,
Jie,
Hello @rfelixmg ,
Thank you for sharing your implementation.
I downloaded your dataset.zip folder and run scripts under src/sota/sun to reproduce your scores on SUN dataset.
The scripts created an experiment folder named: ./experiments/sota/sun/cycle_wgan/0000_TEST_130419_210747 where ls
returns:
-checkpoint
-configuration_130419_210747.json
-git_version.json
-logs
-results
-source
-tensorboard_script.sh
I can share with you the experiment folder, if you need.
According to the contents/hierarchy of the experiment folder, I had to change tester.sh from
#!/usr/bin/env bash
_DBDIR_=./data/
_DATABASE_=sun
_CONFERENCE_=sota
_EPOCH_=epoch_85_85
_MODEL_=./experiments/$_CONFERENCE_/$_DATABASE_/cycle_wgan/classifier/checkpoint/$_EPOCH_/architecture.json
python -m routines.tester -db $_DATABASE_ -rc 1 --load_model $_MODEL_ --dataroot $_DBDIR_
to
#!/usr/bin/env bash
_DBDIR_=./data/
_DATABASE_=sun
_CONFERENCE_=sota
_EPOCH_=epoch_15_15
_MODEL_=./experiments/$_CONFERENCE_/$_DATABASE_/cycle_wgan/0000_TEST_130419_210747/checkpoint/classifier/$_EPOCH_/architecture.json
python -m routines.tester -db $_DATABASE_ -rc 1 --load_model $_MODEL_ --dataroot $_DBDIR_
But then tester.sh gave me this error:
Traceback (most recent call last):
File "/home/X/frwgan/routines/tester.py", line 185, in <module>
model, results = main(options=options, dataset=dataset, knn=knn)
File "/home/X/frwgan/routines/tester.py", line 142, in main
model = load_model(options.load_model, options.architecture.namespace)
File "/home/X/frwgan/routines/aux.py", line 48, in load_model
model = models.__dict__[mtype].__MODEL__()
KeyError: 'rwgan/classifier'
I think this is because "namespace": "rwgan/classifier"
is defined in experiments/sota/sun/cycle_wgan/0000_TEST_130419_210747/checkpoint/classifier/epoch_15_15/architecture.json
. But the weird thing is that running grep -r "rwgan/classifier" ./
in the root repository directory returns only the architecture file created by the experiments ./experiments/sota/sun/cycle_wgan/0000_TEST_130419_210747/checkpoint/classifier/epoch_15_15/architecture.json
. Which means, you never/use 'rwgan/classifier'
in your code.
Could you please help me resolving this issue?
I can't find or install the module named 'util'. Is your another file?
I downloaded all the dataset that is .mat, but how to create the h5?
Best wishes!
Thank you for posting your code online. It is really helpful.
I would like to reproduce your results on FLO, however it appears that the soft links under ./data/flo-reed/
are broken, and refer to missing files:
data.h5 -> data_TENSORFLOW-CVPR_140318_205157.h5
knn.h5 -> FLO_continuous_knn_140318_165701_CVPR.h5
Could you kindly assist with this issue?
Note about a minor issue that I fixed on my own: The script benchmark.sh
looked for the FLO data under ./data/flo
instead of ./data/flo-reed
Many Thanks,
Yuval
Hello @rfelixmg
Thank for providing the Amazing Implementation!!!.
I faced the problem of symlink and solved it by reading #11 .
I am beginner and I have some questions.
Is there anyway to resume epochs from the point where it is stopped?
When I executed, generate_dataset.sh, Output shows, Generating seen and Unseen features and dataset .h5 file is created. How to analyze this H5, How to figure out that unseen feature have been generated, Is there any way to seen them.
Thanks,
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.