Let's say I have a number of videos, and I want to train an SSD/YOLO (or FRCNN) to detect objects. In the case of a large amount of videos, there will be a lot of frames extracted and transferred to images. Can you take, for example, only the fifth frame everytime and thus lower the amount of memory required? If the frames contain a similar info, we can skip some and not impact the results? This is mainly to train faster..

  • $\begingroup$ Theoretically possible, but you have to modify the sorce code at video capture part. I did integrate YOLO with my C++ project and I always skip 5~10 frames per second. But I didn't modify the YOLO source code, because I capture the video frame using OpenCV in my C++ project, not directly from Yolo. $\endgroup$ – gameon67 Mar 21 '19 at 8:27
  • $\begingroup$ Yeah, I think you are right. I used OpenCV and saved the tenth frame, not directly from YOLO as you said. Thanks@! $\endgroup$ – sean2102 Mar 23 '19 at 9:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.