Flow-Guided-Feature-Aggregation
Flow-Guided-Feature-Aggregation copied to clipboard
msracver
The output image is out of order
I run the demo.py. It generates the result.But When I combined all frames into video.It is out of order.Why? Thanks for your help.
not yet
在 2019年10月10日 週四 21:26，blueeda [email protected] 寫道：
Have you solved the problem? I met the same problem.
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/msracver/Flow-Guided-Feature-Aggregation/issues/30?email_source=notifications&email_token=AHRGBDGK2SI2673W2RDBRSDQN7I4PA5CNFSM4FHFBGLKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEA6OH3Y#issuecomment-540861423, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHRGBDDVYAHIFMXJNHVCPA3QN7I4PANCNFSM4FHFBGLA .
Have you solved the problem? I met the same problem.
I have tested on my video with the demo, the accuracy is very low. Did you test your own video?
Have you solved the problem? I met the same problem.
I have tested on my video with the demo, the accuracy is very low. Did you test your own video?
I have tested my own videos, but the accuracy is low. I found that demo.py does not read pictures in order. I think this is the reson of the low accuracy. I am trying to fix this problem.
? but i test other images in the ImageNet, ILSVRC2015_val_00007011 andILSVRC2015_val_00007010， still wrong order but high accuracy. So I think maybe other problem? We fix the problem together, if I fix it I will share .
Have you solved the problem? I met the same problem.
I have tested on my video with the demo, the accuracy is very low. Did you test your own video?
I have tested my own videos, but the accuracy is low. I found that demo.py does not read pictures in order. I think this is the reson of the low accuracy. I am trying to fix this problem.
Hi, did you meet another problem, I can just test not more than about 700 images?
Have you solved the problem? I met the same problem.
I have tested on my video with the demo, the accuracy is very low. Did you test your own video?
I have tested my own videos, but the accuracy is low. I found that demo.py does not read pictures in order. I think this is the reson of the low accuracy. I am trying to fix this problem.
Hi, did you meet another problem, I can just test not more than about 700 images?
Sorry, I have not tested more than 500 pictures. I will try to test more images. We can keep in touch and learn together.
Have you solved the problem? I met the same problem.
I have tested on my video with the demo, the accuracy is very low. Did you test your own video?
I have tested my own videos, but the accuracy is low. I found that demo.py does not read pictures in order. I think this is the reson of the low accuracy. I am trying to fix this problem.
Hi, did you meet another problem, I can just test not more than about 700 images?
I have tested a video with 782 frames and I did not meet any problems.
so I think the order is still not correct. How did you fix the problem. And did you modify it to test real time camera?
so I think the order is still not correct. How did you fix the problem. And did you modify it to test real time camera?
I just changed the order in which the images were read. Demo.py does not read pictures in chronological order, so it can not aggregate features correctly. I have not tested it in real time camera, Now I am reading the codes and try to understand the training and testing process.
I get the right sequence, but accuracy is lower than yolov3, still have many missing detections, have you let it test the video, input the frame one by one ?
I get the right sequence, but accuracy is lower than yolov3, still have many missing detections, have you let it test the video, input the frame one by one ?
I run demo on my video, but I didn't compare the result with other functions.