Update README.md

README.md

@@ -27,10 +27,15 @@ tags:
 
 # TL;DR
 
-Pix2Struct is an image encoder - text decoder model that is trained on image-text pairs for various tasks, including image captionning and visual question answering. The full list of available models can be found on the Table 1 of the paper:
+## Details for Pix2Struct-RefExp: (Based on their [pre-processing](https://github.com/google-research/pix2struct/blob/main/pix2struct/preprocessing/convert_refexp.py))
 
-![Table 1 - paper](https://s3.amazonaws.com/moonup/production/uploads/1678712985040-62441d1d9fdefb55a0b7d12c.png)
+-> __Input__: An image with a bounding box drawn on it around a candidate object and a header containing the referring expression (stored in the image feature).
+
+-> __Output__: A boolean flag (parse feature) indicating whether the candidate object is the correct referent of the referring expression.
 
+__Pix2Struct__ is an image encoder - text decoder model that is trained on image-text pairs for various tasks, including image captionning and visual question answering. The full list of available models can be found on the Table 1 of the paper:
+
+![Table 1 - paper](https://s3.amazonaws.com/moonup/production/uploads/1678712985040-62441d1d9fdefb55a0b7d12c.png)
 
 The abstract of the model states that:
 > Visually-situated language is ubiquitous—sources range from textbooks with diagrams to web pages with images and tables, to mobile apps with buttons and