Input error for CV inference API via Postman

Hi all,

I’m receiving the following error:
"Error in `inputs`: Invalid image: b'------WebKitFormBoun..'"

When I try and use the inference API on an Image Segmentation model (CIDAS/clipseg-rd64-refined).

I’ve tried send the image as a file (JPEG and PNG formats) and via a URL.

Point of note:

I discovered the CV model (CIDAS/clipseg-rd64-refined) doesn’t work with the API, but I get the same error with the model (facebook/mask2former-swin-large-cityscapes-semantic · Hugging Face) which does with the API.