Laughing

ok after checked out the config files, from the name of `yolo_world_l_dual` and `yolo_world_l_t2i`, I guess the former is using both `Text-guided CSPLayers` and `Image-Pooling`, and the latter only uses...

Another question, I noticed there's an empty placeholder `" "` manually be added here: https://github.com/AILab-CVC/YOLO-World/blob/7d43247a0e67af0858f63fdc0ec0a4a9fe0a79b4/demo.py#L49 which I think it's probably a special trick to be the `background` prompt for open-vocabulary...

@wondervictor Awesome! Thanks for the information!

@Baboom-l yeah I was thinking about the same thing, but haven't really looked through the code of training process.

@wondervictor Cool!

@glenn-jocher @farhanrw Guys the final augmented oriented bounding boxes are determined by the algorithm `cv2.minAreaRect()` that we uses in `xyxyxyxy2xywhr` function. https://github.com/ultralytics/ultralytics/blob/5a82b511074617b9daa25000e5672c5d56188374/ultralytics/utils/ops.py#L539 For objects that are cut-off a lot by...

@glenn-jocher Done in https://github.com/ultralytics/ultralytics/pull/11653/commits/87f46e77128af9125b9da30033ed3c8ad56d3f1e. Currently classify model would only output `names, confidence, class` keys(no box), and obb by default returns `xyxyxyxy` format for `box` with keys `x1, y1, x2, y2,...

@Y-T-G @Burhan-Q Hey guys! Nice catch! yes it's true the regressed bounding boxes are limited by the `reg_max` value. Sorry for the delay though, I had to test it locally...

@PallottaEnrico @glenn-jocher yeah allowing users to modify this arg could be very helpful. But I just don't know how much effect modifying this arg will have on users' training. I...

@PallottaEnrico ok so you did get much better results after modifying this arg, then I think it's worth to add it :). @glenn-jocher