---
license: mit
license_link: https://huggingface.co/microsoft/Florence-2-base-ft/resolve/main/LICENSE
pipeline_tag: image-text-to-text
tags:
- vision
- ocr
- segmentation
datasets:
- yifeihu/TF-ID-arxiv-papers
---
# TF-ID: Table/Figure IDentifier for academic papers

## Model Summary

TF-ID (Table/Figure IDentifier) is a family of object detection models finetuned to extract tables and figures in academic papers created by [Yifei Hu](https://x.com/hu_yifei). They come in four versions:
| Model   | Model size | Model Description | 
| ------- | ------------- |   ------------- |  
| TF-ID-base[[HF]](https://huggingface.co/yifeihu/TF-ID-base) | 0.23B  | Extract tables/figures and their caption text  
| TF-ID-large[[HF]](https://huggingface.co/yifeihu/TF-ID-large) (Recommended) | 0.77B  | Extract tables/figures and their caption text  
| TF-ID-base-no-caption[[HF]](https://huggingface.co/yifeihu/TF-ID-base-no-caption) | 0.23B  | Extract tables/figures without caption text
| TF-ID-large-no-caption[[HF]](https://huggingface.co/yifeihu/TF-ID-large-no-caption) (Recommended) | 0.77B  | Extract tables/figures without caption text
All TF-ID models are finetuned from [microsoft/Florence-2](https://huggingface.co/microsoft/Florence-2-large-ft) checkpoints.

- The models were finetuned with papers from Hugging Face Daily Papers. All bounding boxes are manually annotated and checked by humans.
- TF-ID models take an image of a single paper page as the input, and return bounding boxes for all tables and figures in the given page. 
- TF-ID-base and TF-ID-large draw bounding boxes around tables/figures and their caption text.
- TF-ID-base-no-caption and TF-ID-large-no-caption draw bounding boxes around tables/figures without their caption text.

**Large models are always recommended!**

![image/png](https://huggingface.co/yifeihu/TF-ID-base/resolve/main/td-id-caption.png)

Object Detection results format: 
{'\<OD>': {'bboxes': [[x1, y1, x2, y2], ...], 
'labels': ['label1', 'label2', ...]} }

## Training Code and Dataset
- Dataset: [yifeihu/TF-ID-arxiv-papers](https://huggingface.co/datasets/yifeihu/TF-ID-arxiv-papers)
- Code: [github.com/ai8hyf/TF-ID](https://github.com/ai8hyf/TF-ID)

## Benchmarks

We tested the models on paper pages outside the training dataset. The papers are a subset of huggingface daily paper.

Correct output - the model draws correct bounding boxes for every table/figure in the given page.

| Model                                                         | Total Images | Correct Output | Success Rate |
|---------------------------------------------------------------|--------------|----------------|--------------|
| TF-ID-base[[HF]](https://huggingface.co/yifeihu/TF-ID-base)   | 258          | 251            | 97.29%       |
| TF-ID-large[[HF]](https://huggingface.co/yifeihu/TF-ID-large) | 258          | 253            | 98.06%       |

| Model                                                         | Total Images | Correct Output | Success Rate |
|---------------------------------------------------------------|--------------|----------------|--------------|
| TF-ID-base-no-caption[[HF]](https://huggingface.co/yifeihu/TF-ID-base-no-caption)   | 261          | 253            | 96.93%       |
| TF-ID-large-no-caption[[HF]](https://huggingface.co/yifeihu/TF-ID-large-no-caption) | 261          | 254            | 97.32%       |

Depending on the use cases, some "incorrect" output could be totally usable. For example, the model draw two bounding boxes for one figure with two child components.
 
## How to Get Started with the Model

Use the code below to get started with the model.

```python
import requests
from PIL import Image
from transformers import AutoProcessor, AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained("yifeihu/TF-ID-base", trust_remote_code=True)
processor = AutoProcessor.from_pretrained("yifeihu/TF-ID-base", trust_remote_code=True)

prompt = "<OD>"
url = "https://huggingface.co/yifeihu/TF-ID-base/resolve/main/arxiv_2305_10853_5.png?download=true"
image = Image.open(requests.get(url, stream=True).raw)

inputs = processor(text=prompt, images=image, return_tensors="pt")
generated_ids = model.generate(
    input_ids=inputs["input_ids"],
    pixel_values=inputs["pixel_values"],
    max_new_tokens=1024,
    do_sample=False,
    num_beams=3
)

generated_text = processor.batch_decode(generated_ids, skip_special_tokens=False)[0]
parsed_answer = processor.post_process_generation(generated_text, task="<OD>", image_size=(image.width, image.height))
print(parsed_answer)
```

To visualize the results, see [this tutorial notebook](https://colab.research.google.com/github/roboflow-ai/notebooks/blob/main/notebooks/how-to-finetune-florence-2-on-detection-dataset.ipynb) for more details.

## BibTex and citation info

```
@misc{TF-ID,
  author = {Yifei Hu},
  title = {TF-ID: Table/Figure IDentifier for academic papers},
  year = {2024},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/ai8hyf/TF-ID}},
}
```