I am trying to implement the original YOLO architecture for object detection, but I am using the COCO dataset. However, I am a bit confused about the image sizes of COCO. The original YOLO was trained on the VOC dataset and it is designed to take 448x448 size images. Since I am using COCO, I thought of cropping down the images to that size. But that would mean I would have to change the annotations file and it might make the process of object detection a bit harder because some of the objects might not be visible.

I am pretty new to this, so I am not sure if this is the right way or what are some other things that I can do. Any help will be appreciated.

  • $\begingroup$ What Yolo version, v1, v2 or v3? As you know there are 2 iterations over the original paper. What version are you trying to use? If you specify it, I might be able to help. $\endgroup$
    – JVGD
    Feb 5 '20 at 16:59

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.