peiwang062

> > Hi, thanks for sharing the great work! But why not opening issues? But anyway, I tried to reproduce the ferret-7b evaluation results on refcocog, but my reproduced results...

> > > > Hi, thanks for sharing the great work! But why not opening issues? But anyway, I tried to reproduce the ferret-7b evaluation results on refcocog, but my...

The current code only support batch size=1. This if statement is from original [deformableDETR](https://github.com/fundamentalvision/Deformable-DETR/blob/11169a60c33333af00a4849f1808023eba96a931/util/misc.py#L319) implementation, we didn't make any change on it.

The segmentation mask is not a must. On COCO, we basically followed the setting of point DETR paper, '(a) if the object has instance segmentation, randomly sample a point from...

tensor_label_k is the weakly augmented image. We just followed the standard teacher-student framework. It is just one image, shouldn't be too big.

Bounding boxes still exist but we didn't use them on your two cases. Please see Line147-184 of engine.py. We route each examples according to their label types. If you only...

Our code only supports two options for the resolution (600/800). Other settings may cause unexpected errors. After burnin, there will be a model duplication. This doubles the memory. So maybe...

I think a simplest way is to organize your customized data and its annotation to coco format. For fully labeled data, we still need its bbox, so can't let annotation...

Hi, any change made on the code? for example, batch size. The current code only supports batch size equal to 1.