Fix the bug by latest transformers and update req.txt.

README.md

@@ -13,7 +13,7 @@ pipeline_tag: image-segmentation
 license: mit
 ---
 
-> This BiRefNet was trained with images in `512x512` for higher resolution inference.
+> This BiRefNet was trained with images in `512x512` for faster and more accurate lower resolution inference.
 ### Performance:
 > All tested in FP16 mode.