yanbeic / val Goto Github PK
View Code? Open in Web Editor NEWTensorflow Implementation on Paper [CVPR2020]Image Search with Text Feedback by Visiolinguistic Attention Learning
License: Apache License 2.0
Tensorflow Implementation on Paper [CVPR2020]Image Search with Text Feedback by Visiolinguistic Attention Learning
License: Apache License 2.0
Dear everyone who's working in this field,
I am reimplementing this work in Pytorch. My code doesn't give the same result as the reported results in the paper. Can you guys help me out by seeing my code? I will appreciate it very much.
Hi, I know this is a bit more to ask, but just in case- is there a pyTorch implementation available, or by any chance an easy transfer of model to the pyTorch?
Thanks.
I am running python generate_groundtruth.py --dataset='shoes' --data_path=''
and falling to this error.
FileNotFoundError: [Errno 2] No such file or directory: 'datasets/shoes/shoes-tag-test.txt'
Hello,
some url links are broken in fashioniq. How can I get the right pictures?
Hi, there.
Thank you for the great work!
According to fashion200k repo, there are two versions of Fashion200k pictures, namely cropped detected images and original images, which one do you use in your implementation?
I have some problems. Please help me.
bash scripts/run_fashion_iq.sh
and falling to this error.Traceback (most recent call last):
File "test_val.py", line 220, in <module>
main()
File "test_val.py", line 48, in main
query_images = np.load(filename)
File "/home/huynhtruc0309/anaconda3/envs/truchlp/lib/python3.7/site-packages/numpy/lib/npyio.py", line 416, in load
fid = stack.enter_context(open(os_fspath(file), "rb"))
FileNotFoundError: [Errno 2] No such file or directory: 'save_model/fashion_iq/val_resnet_v2_50_ml/query_images.npy'
Hello, I downloaded the Google cloud disk dataset (fashion-200k) according to your connection, but I haven't passed your download license. Can you pass my application?
I got this errors when I try to run " python generate_groundtruth.py --dataset='fashion_iq' --data_path='fashion_iq' " and " python read_glove.py --dataset='fashion_iq' --data_path='datasets/fashion_iq/image_data' " :
...
file = open(filename, "r")
FileNotFoundError: [Errno 2] No such file or directory: 'datasets/fashion_iq/tags/asin2attr-test.txt'
I found the .txt in ‘datasets/fashion_iq/tags/’ should be generated by " python generate_tags.py --dataset='fashion_iq' ", but the code shows no "asin2attr-test.txt" is generated. It only generated 9 .txt as follows:
writepaths = [
'datasets/fashion_iq/tags/asin2attr-train-dress.txt',
'datasets/fashion_iq/tags/asin2attr-train-shirt.txt',
'datasets/fashion_iq/tags/asin2attr-train-toptee.txt',
'datasets/fashion_iq/tags/asin2attr-val-dress.txt',
'datasets/fashion_iq/tags/asin2attr-val-shirt.txt',
'datasets/fashion_iq/tags/asin2attr-val-toptee.txt',
'datasets/fashion_iq/tags/asin2attr-test-dress.txt',
'datasets/fashion_iq/tags/asin2attr-test-shirt.txt',
'datasets/fashion_iq/tags/asin2attr-test-toptee.txt',
]
Is there any wrong in your codes?
I got this errors when I try to run python generate_caption_pairs.py --dataset='fashion_iq'
Traceback (most recent call last):
File "generate_tags.py", line 10, in
'dataset', "fashion_iq or shoes")
File "C:\ProgramData\Miniconda3\lib\site-packages\tensorflow_core\python\platform\flags.py", line 58, in wrapper
return original_function(*args, **kwargs)
TypeError: DEFINE_string() missing 1 required positional argument: 'help'
When I run "python test_val.py --feature_dir=${FEAT_DIR} --batch_size=20 --dataset=fashoin_iq --subset=toptee, I got error at Traceback (most recent call last):
File "test_val.py", line 218, in
main()
File "test_val.py", line 179, in main
gt_label = order[gt_mask[i, :]]
IndexError: index 3920 is out of bounds for axis 0 with size 3920
and when the subset is "shirt", the Indexerror is: IndexError: boolean index did not match indexed array along dimension 0; dimension is 2902 but corresponding boolean dimension is 3089
when the subset is "dress", the Indexerror is: boolean index did not match indexed array along dimension 0; dimension is 2902 but corresponding boolean dimension is 2628
In fashion-iq dataset, there are two captions available, in paper "The fashion iq dataset: Retrieving images by combining side information and relative natural language feedback", they concatenate two captions with special symbol "<and>" and in doing so they treat two captions as one. So each ref-cap-tgt pair in json file only results in one pair.
For more detail, see their released source code
However, in your evaluation process, the file "fashion_iq-val-cap.txt" shows that you threat them as two individual pairs. If the results in your paper follow the same evaluation process, I doubt your results are not comparable with early published work.
When running run_fashion_iq.sh
to train_val.py I got to this error
scripts/run_fashion_iq.sh: line 44: 1876 Segmentation fault (core dumped) python train_val.py --checkpoint_dir=${STAGE1_DIR} --pretrain_checkpoint_dir=${PRETRAIN_DIR} --image_model=${CNN} --data_path=${DATA_DIR} --batch_size=32 --print_span=20 --save_length=10000 --text_model=${TEXT_MODEL} --joint_embedding_size=${JOINT_SIZE} --word_embedding_size=${WORD_SIZE} --text_embedding_size=${TEXT_SIZE} --margin=0.2 --text_projection_dropout=0.9 --init_learning_rate=0.0002 --dataset=${DATASET} --train_length=50000 --constant_lr=True --image_size=${IMG_SIZE} --image_feature_name='before_pool' --augmentation=${AUGMENT} --word_embedding_dir=${PRE_TEXT}
How can Segmentation happen? Please help me, thank you.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 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. 📊📈🎉
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 ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.