Grobid Vs Nougat

[mv96]

Hello,

I was recently reading the NOUGAT paper (https://arxiv.org/pdf/2308.13418.pdf) and one of the tables of the paper directly compares the performance with Grobid see the image below👇

from the image above it is showing significant gains over GROBID on the task of formula identification which could possibly cover the problem discussed in this thread #825 .

As a small experiment I tried to give the same pdf to an open implementation of Nougat,

click here to see the hosted Hugging face space👇 https://huggingface.co/spaces/ysharma/nougat

And I can see that Nougat model works decent in this case

I was using a task where I require the segmentation of paragraphs (text cut in blocks) instead of plain text output, so I was wondering if there was a possibility of using Nougat backbone with Grobid ?

[lfoppiano]

I think, perhaps, one important information missing from that table is the runtime, which is mentioned at the end of the paper:

The principle with Grobid is to avoid rasterisation / OCR processing that are much slower (several orders of magnitude).

For paragraphs, I think Grobid works well given the small dataset used for training the fulltext model, however regarding the formulas, supposely you can get the coordinates of the formulas, cut it from the PDF document, and pass it to any postprocess you can use (nougat, or other)