Vision-Language-Transformer icon indicating copy to clipboard operation
Vision-Language-Transformer copied to clipboard

Test with own dataset

Open wooj0216 opened this issue 1 year ago • 2 comments

Thank you for your great work!

I am curious about testing with my own dataset. I just want to get a mask image from my data.

Thank you very much.

wooj0216 avatar May 08 '24 14:05 wooj0216

Thank you for your great work!

I am curious about testing with my own dataset. I just want to get a mask image from my data.

Thank you very much.

Thank you for your interest to our work. You can download our trained models and run the inference code python vlt.py test [PATH_TO_CONFIG_FILE], input your image and a sentence describing the target object.

You could also considering our recent method ReLA.

henghuiding avatar May 09 '24 05:05 henghuiding

Thank you for the response.

I am little confused about preparing the dataset json file. Can you give me an example json file or samples for inference?

Also, I would like to ask you if I can use images with various shapes such as (192, 384) or (120, 240).

I really appreciate for your help.

wooj0216 avatar May 09 '24 06:05 wooj0216