I consider pre-training a YOLOv5 with Google Open Images Object Detection dataset. The dataset includes general domain categories with ~15 M box samples. After the pre-training is done, I will fine-tune the model on MS COCO dataset.

I would like to do it, if I can improve AP by ~7%. Do you think that it is possible, and I have a logical expectation? Unfortunately, I could not find anywhere anyone has tried an Open Images pre-trained object detector with MSCOCO training.

  • 1
    $\begingroup$ To clarify, by MSCOCO, you mean Microsoft's COCO dataset, right? $\endgroup$
    – nbro
    Nov 19, 2020 at 12:39
  • $\begingroup$ yes, that was right $\endgroup$
    – yusuf
    Nov 21, 2020 at 17:55


You must log in to answer this question.

Browse other questions tagged .