search

ziqi-jin / finetune-anything

Fine-tune SAM (Segment Anything Model) for computer vision tasks such as semantic segmentation, matting, detection ... in specific scenarios
MIT License
766 stars 55 forks source link

Why set the mask_scale=1 #33

Open luowei0701 opened 1 year ago

luowei0701 commented 1 year ago

An interesting job, but I have some questions. Why set mask_scale=1 with multimask_output=True instead of mask_scale=0 with multimask_output=False here? https://github.com/ziqi-jin/finetune-anything/blob/85a0658cb0011a504aa73f2399cf4ab305cc8a65/extend_sam/mask_decoder_heads.py#L192

ziqi-jin commented 1 year ago

SAM supports three different scale instance mask outputs for each input default, Users could choose the scale of mask they want by setting the mask_scale, you are very patient, I forgot to add this into the tutorial.