search

fu123456 / SHIQ

Project for CVPR21 paper: "A Multi-Task Network for Joint Specular Highlight Detection and Removal".
85 stars 16 forks source link

Can you provide test results? #4

Closed nachifur closed 2 years ago

nachifur commented 2 years ago

We noticed that you posted a test model. Before deciding whether to go further, the researcher seems to want to see the test image directly. Can you provide test results? For example, a zip file containing test images (.png) available for download. Looking forward to your reply.

fu123456 commented 2 years ago

It will come soon!

nachifur commented 2 years ago

If you can improve this project as soon as possible, we will be very grateful to you. It is of great significance to the community.

fu123456 commented 2 years ago

Please use this bash file https://github.com/fu123456/SHIQ/blob/main/network/test_list.sh to generate results.

fu123456 commented 2 years ago

also see https://github.com/fu123456/SHIQ/blob/main/network/generate_test_list.sh

nachifur commented 2 years ago

Thank you very much for your reply. Which training data set corresponds to the test model you provided? Can I directly use the model to test on SHIQ, CLH, LIME? image

The problem still exists.

Please use this bash file https://github.com/fu123456/SHIQ/blob/main/network/test_list.sh to generate results. also see https://github.com/fu123456/SHIQ/blob/main/network/generate_test_list.sh

Your answer does not seem to solve the problem.

nachifur commented 2 years ago

We want to make a visual comparison with your results on SHIQ, CLH and LIME. Can you provide useful guidance?

fu123456 commented 2 years ago

Our training model corresponds to our SHIQ dataset, without relating to CLH and LIME datasets. Please contact the author of LIME dataset to obtain this dataset. If you test LIME dataset, we need to retrain the network and then test.

nachifur commented 2 years ago

@fu123456 I have also observed the following error, which based on the test process in readme. image

This error was raised in the following issue and has not been effectively resolved. https://github.com/fu123456/SHIQ/issues/3#issuecomment-988422829

fu123456 commented 2 years ago

@nachifur this problem has been solved. See https://github.com/fu123456/SHIQ/issues/3