search

JXingZhao / ContrastPrior

The Code of Contrast Prior and Fluid Pyramid Integration for RGBD Salient Object Detection(CVPR2019)
51 stars 8 forks source link

About the test script #2

Open zhangqiudan opened 5 years ago

zhangqiudan commented 5 years ago

Dear author, Thank you very much for your public code. I am confused about your test script.

  1. About the test.lst; Does the test.lst contain the path of the left image and the depth image? ./nju2k/LR/000799_left.png ./depth/000799_left.jpg

  2. About the test.py; I didn't see the input of depth information from this test.py script. And
    does the 'sal_lst' mean the final predicted salient object map? What does the 'crf_lst' mean?

    Thank you so much.

JXingZhao commented 5 years ago

Dear sir/madam:

  1. The test.lst only contains the left image, like this: ./RGB/RGBD_data_100.jpg. However, if u set the test.lst as ./nju2k/LR/000799_left.png ./depth/000799_left.jpg. It is also okay.
  2. sal_lst means the list of final predicted salient object map. crf_lst means the list of the result after CRF. In the presented results, we don't use the CRF, so u can ignore it. The depth list is generated in the data layer. The details are showed in /caffe/lib/ImageLabelDataTest.py. In this file, you can figure out how we generate the test data. .

发件人：zhangqiudan notifications@github.com 发送日期：2019-08-05 12:01:23 收件人：JXingZhao/ContrastPrior ContrastPrior@noreply.github.com 抄送人：Subscribed subscribed@noreply.github.com 主题：[JXingZhao/ContrastPrior] About the test script (#2) Dear author, Thank you very much for your public code. I am confused about your test script.

  1. About the test.lst; Does the test.lst contain the path of the left image and the depth image? ./nju2k/LR/000799_left.png ./depth/000799_left.jpg
  2. About the test.py; I didn't see the input of depth information from this test.py script. And does the 'sal_lst' mean the final predicted salient object map? What does the 'crf_lst' mean? Thank you so much. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.
zhangqiudan commented 5 years ago

Thank you very much for your reply.

JXingZhao notifications@github.com 于2019年8月5日周一 下午9:05写道：

Dear sir/madam:

  1. The test.lst only contains the left image, like this: ./RGB/RGBD_data_100.jpg. However, if u set the test.lst as ./nju2k/LR/000799_left.png ./depth/000799_left.jpg. It is also okay.
  2. sal_lst means the list of final predicted salient object map. crf_lst means the list of the result after CRF. In the presented results, we don't use the CRF, so u can ignore it. The depth list is generated in the data layer. The details are showed in /caffe/lib/ImageLabelDataTest.py. In this file, you can figure out how we generate the test data. .

发件人：zhangqiudan notifications@github.com 发送日期：2019-08-05 12:01:23 收件人：JXingZhao/ContrastPrior ContrastPrior@noreply.github.com 抄送人：Subscribed subscribed@noreply.github.com 主题：[JXingZhao/ContrastPrior] About the test script (#2) Dear author, Thank you very much for your public code. I am confused about your test script.

  1. About the test.lst; Does the test.lst contain the path of the left image and the depth image? ./nju2k/LR/000799_left.png ./depth/000799_left.jpg
  2. About the test.py; I didn't see the input of depth information from this test.py script. And does the 'sal_lst' mean the final predicted salient object map? What does the 'crf_lst' mean? Thank you so much. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/JXingZhao/ContrastPrior/issues/2?email_source=notifications&email_token=AFNOWEG6HLRRKMIG3XTV4GDQDAQPZA5CNFSM4IJHIKJ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD3RYAFI#issuecomment-518225941, or mute the thread https://github.com/notifications/unsubscribe-auth/AFNOWEEYVTJVORHKQFJXWY3QDAQPZANCNFSM4IJHIKJQ .

zhangqiudan commented 5 years ago

Dear author, When I use the provided final.caffemodel to predict the saliency map, the output is a black image. The printed 'res' is all zeros.

JXingZhao commented 5 years ago

I suggest that you can debug step by step. First, you can check whether the picture was successfully input into the network, and then check whether the input picture is correct.

发件人：zhangqiudan notifications@github.com 发送日期：2019-08-08 17:16:59 收件人：JXingZhao/ContrastPrior ContrastPrior@noreply.github.com 抄送人：JXingZhao zhaojiaxing@mail.nankai.edu.cn,Comment comment@noreply.github.com 主题：Re: [JXingZhao/ContrastPrior] About the test script (#2) Dear author, When I use the provided final.caffemodel to predict the saliency map, the output is a black image. The printed 'res' is all zeros. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

JXingZhao commented 5 years ago

Another possible reason is that the input results of the network are not scaled to 0-255.

发件人：zhangqiudan notifications@github.com 发送日期：2019-08-08 17:16:59 收件人：JXingZhao/ContrastPrior ContrastPrior@noreply.github.com 抄送人：JXingZhao zhaojiaxing@mail.nankai.edu.cn,Comment comment@noreply.github.com 主题：Re: [JXingZhao/ContrastPrior] About the test script (#2) Dear author, When I use the provided final.caffemodel to predict the saliency map, the output is a black image. The printed 'res' is all zeros. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

zhangqiudan commented 5 years ago

Thank you very much for your reply, I will try to debug the program.

JXingZhao notifications@github.com 于2019年8月9日周五 下午2:55写道：

Another possible reason is that the input results of the network are not scaled to 0-255.

发件人：zhangqiudan notifications@github.com 发送日期：2019-08-08 17:16:59 收件人：JXingZhao/ContrastPrior ContrastPrior@noreply.github.com 抄送人：JXingZhao zhaojiaxing@mail.nankai.edu.cn,Comment < comment@noreply.github.com> 主题：Re: [JXingZhao/ContrastPrior] About the test script (#2) Dear author, When I use the provided final.caffemodel to predict the saliency map, the output is a black image. The printed 'res' is all zeros. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/JXingZhao/ContrastPrior/issues/2?email_source=notifications&email_token=AFNOWECTRQWMO5T64V7BMGLQDUIHPA5CNFSM4IJHIKJ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD35ZA7I#issuecomment-519803005, or mute the thread https://github.com/notifications/unsubscribe-auth/AFNOWEBALW7NNPIC247ZHYTQDUIHPANCNFSM4IJHIKJQ .