EOFerror when python Random_clip_valid.py

[cckenny]

2%|▏ | 29/1261 [00:08<06:00, 3.41it/s]test acc: 0.8620689655172413 test_correct_num: 75 total_v: 87.0 Loading a video clip from F:/UCF-101//Archery/v_Archery_g06_c06... Loading a video clip from F:/UCF-101//Archery/v_Archery_g07_c04...

Traceback (most recent call last): File "E:/graduate/action-recognition/code/Random_clip_valid.py", line 197, in modelpath="./") File "E:/graduate/action-recognition/code/Random_clip_valid.py", line 184, in test shuffle=False File "E:\graduate\action-recognition\code\input_data_v1.py", line 97, in read_clip_and_label raise EOFError EOFError

Process finished with exit code 1

Thanks

[cckenny]

I passed it but am I right？

[cckenny]

what's more i wonder why my accuracy is always 1.0 during the whole training? the predict result is right i guess At least they are between 0 and 1.0

[binhhoangtieu]

This error may be caused by the number of extracted file is less than 16 (look into this folder F:/UCF-101//Archery/v_Archery_g07_c04)

[cckenny]

thanks，u r right.but why? i used the shell code in ubuntu to extract those frames. do you know why my accuracy is always 1.0 during the whole training?

[cckenny]

I saw your comment I think our accuracy may have the same bug.

[binhhoangtieu]

thanks，u r right.but why? i used the shell code in ubuntu to extract those frames.

Because your video is not long enough, you can change the file "convert_video_to_images.sh" for extracting more frames

do you know why my accuracy is always 1.0 during the whole training?

I'm not sure but you should check your downloaded files (models, train.list and test.list), you should modify this path before training (do not use original train.list and test.list)

Hope this helps

[ZhiQue930716]

@cckenny i met the same error,how did you fixed this ?

[PBRAOS]

there is a simple fix. just comment the Error exceptions.

[Luhn-Saul]

thanks，u r right.but why? i used the shell code in ubuntu to extract those frames.

Because your video is not long enough, you can change the file "convert_video_to_images.sh" for extracting more frames

do you know why my accuracy is always 1.0 during the whole training?

I'm not sure but you should check your downloaded files (models, train.list and test.list), you should modify this path before training (do not use original train.list and test.list)

Hope this helps

Many of the extracted documents are less than 16(eg.:Archery/v_Archery_g05_c04,Archery/v_Archery_g07_c02&c03&c04&c05&c06,and Archery/v_Archery_g17_c04,and BalanceBeam/v_BalanceBeam_g02_c03,and BalanceBeam/v_BalanceBeam_g03_c04,and BalanceBeam/v_BalanceBeam_g10_c02,and BalanceBeam/v_BalanceBeam_g11_c02,and BalanceBeam/v_BalanceBeam_g12_c01&02,and BalanceBeam/v_BalanceBeam_g13_c04,ect.) Can changing the file "convert_video_to_images.sh" really solve the problem? Thank you and I will be looking forward to your reply.

[cckenny]

yeap

Luhn-Saul [email protected] 于2019年2月17日周日 下午1:53写道：

thanks，u r right.but why? i used the shell code in ubuntu to extract those frames.

Because your video is not long enough, you can change the file "convert_video_to_images.sh" for extracting more frames

do you know why my accuracy is always 1.0 during the whole training?

I'm not sure but you should check your downloaded files (models, train.list and test.list), you should modify this path before training (do not use original train.list and test.list)

Hope this helps

Many of the extracted documents are less than 16(eg.:Archery/v_Archery_g05_c04,Archery/v_Archery_g07_c02&c03&c04&c05&c06,and Archery/v_Archery_g17_c04,and BalanceBeam/v_BalanceBeam_g02_c03,and BalanceBeam/v_BalanceBeam_g03_c04,and BalanceBeam/v_BalanceBeam_g10_c02,and BalanceBeam/v_BalanceBeam_g11_c02,and BalanceBeam/v_BalanceBeam_g12_c01&02,and BalanceBeam/v_BalanceBeam_g13_c04,ect.) Can changing the file "convert_video_to_images.sh" really solve the problem? Thank you and I will be looking forward to your reply.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/hx173149/C3D-tensorflow/issues/53#issuecomment-464420377, or mute the thread https://github.com/notifications/unsubscribe-auth/AjyyQ2yeS5LJUMOj6hsWSqrR4WJr8i3xks5vOO5xgaJpZM4TgGH2 .

[Luhn-Saul]

All right. Could you tell me detailedly how you changing it? I hope to get your help! Thanks.