Skip to content

Revert "project page" #3

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
AIDAS-Lab wants to merge 1 commit into from
Closed

Revert "project page" #3

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
151 changes: 146 additions & 5 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,148 @@
# AIDAS Lab Project Page Template
# MathSpeech: Leveraging Small LMs for Accurate Conversion in Mathematical Speech-to-Formula

This repository hosts the source code for the AIDAS Lab Project Page.
The template was originally derived from [Nerfies](https://nerfies.github.io).
## Abstract
In various academic and professional settings, such as mathematics lectures or research presentations, it is often necessary to convey mathematical expressions orally. However, reading mathematical expressions aloud without accompanying visuals can significantly hinder comprehension, especially for those who are hearing-impaired or rely on subtitles due to language barriers. For instance, when a presenter reads Euler's Formula, current Automatic Speech Recognition (ASR) models often produce a verbose and error-prone textual description (e.g., e to the power of i x equals cosine of x plus i $\textit{side}$ of x), instead of the concise LaTeX format (i.e., $e^{ix} = \cos(x) + i\sin(x)$), which hampers clear understanding and communication. To address this issue, we introduce MathSpeech, a novel pipeline that integrates ASR models with small Language Models (sLMs) to correct errors in mathematical expressions and accurately convert spoken expressions into structured LaTeX representations. Evaluated on a new dataset derived from lecture recordings, MathSpeech demonstrates LaTeX generation capabilities comparable to leading commercial Large Language Models (LLMs), while leveraging fine-tuned small language models of only 120M parameters.
Specifically, in terms of CER, BLEU, and ROUGE scores for LaTeX translation, MathSpeech demonstrated significantly superior capabilities compared to GPT-4o. We observed a decrease in CER from 0.390 to 0.298, and higher ROUGE/BLEU scores compared to GPT-4o.

# Website License
<a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/"><img alt="Creative Commons License" style="border-width:0" src="https://i.creativecommons.org/l/by-sa/4.0/88x31.png" /></a><br />This work is licensed under a <a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/">Creative Commons Attribution-ShareAlike 4.0 International License</a>.
### This study is accepted for the AAAI-25 Main Technical Track.

Here, you can find the benchmark dataset, experimental code, and fine-tuned model checkpoints for MathSpeech, which we have developed for our research.

If you want to view the detailed information about the dataset used in this study or additional experimental results such as latency measurements included in the appendix, please refer to the version uploaded on arXiv.

---

## Benchmart Dataset
The MathSpeech benchmark dataset is available on [huggingface🤗](https://huggingface.co/datasets/AAAI2025/MathSpeech) or through the following [link](https://drive.google.com/drive/folders/1M8_IVcesO2EwNcl9zwxY6UgqAmSODzgq?usp=sharing).

- [MathSpeech in huggingface🤗 dataset](https://huggingface.co/datasets/AAAI2025/MathSpeech)
- [Google Drive link for dataset](https://drive.google.com/drive/folders/1M8_IVcesO2EwNcl9zwxY6UgqAmSODzgq?usp=sharing)


#### Dataset statistics
<table border="1" style="border-collapse: collapse; width: 50%;">
<thead>
<tr>
<th style="text-align: left;">The number of files</th>
<td>1,101</td>
</tr>
</thead>
<thead>
<tr>
<th style="text-align: left;">Total Duration</th>
<td>5583.2 seconds</td>
</tr>
</thead>
<tbody>
<tr>
<th style="text-align: left;">Average Duration per file</th>
<td>5.07 seconds</td>
</tr>
<tr>
<th style="text-align: left;">The number of speakers</th>
<td>10</td>
</tr>
<tr>
<th style="text-align: left;">The number of men</th>
<td>8</td>
</tr>
<tr>
<th style="text-align: left;">The number of women</th>
<td>2</td>
</tr>
<tr>
<th style="text-align: left;">source</th>
<td><a href="https://www.youtube.com/@mitocw" target="_blank">[MIT OpenCourseWare]</td>
</tr>
</tbody>
</table>



#### WERs of various ASR models on the Mathspeech benchmark
<table style="width:100%; border-collapse: collapse;">
<thead>
<tr>
<th></th>
<th>Models</th>
<th>Params</th>
<th>WER(%) (Leaderboard)</th>
<th>WER(%) (Formula)</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="4">OpenAI</td>
<td>Whisper-base</td>
<td>74M</td>
<td>10.3</td>
<td>34.7</td>
</tr>
<tr>
<td>Whisper-small</td>
<td>244M</td>
<td>8.59</td>
<td>29.5</td>
</tr>
<tr>
<td>Whisper-largeV2</td>
<td>1550M</td>
<td>7.83</td>
<td>31.0</td>
</tr>
<tr>
<td>Whisper-largeV3</td>
<td>1550M</td>
<td>7.44</td>
<td>33.3</td>
</tr>
<tr>
<td>NVIDIA</td>
<td>Canary-1B</td>
<td>1B</td>
<td>6.5</td>
<td>35.2</td>
</tr>
</tbody>
</table>

##### The WER for Leaderboard was from the [HuggingFace Open ASR Leaderboard](https://huggingface.co/spaces/hf-audio/open_asr_leaderboard), while the WER for Formula was measured using our MathSpeech Benchmark. This value is based on results as of 2024-08-16.


---
## MathSpeech Checkpoint
You can download the MathSpeech checkpoint from the following [link](https://drive.google.com/file/d/1m0cCpDDkOb7FltjLPVlg4ZCZSSSWZgS2/view?usp=sharing).

### Experiments codes

You can find the MathSpeech evaluation code, and the prompts used for the LLMs in the experiments at the following [link](https://github.com/hyeonsieun/MathSpeech/tree/main/Experiments).

### Ablation Study codes

You can find the code used in our Ablation Study at the following [link](https://github.com/hyeonsieun/MathSpeech/tree/main/Ablation_Study).

---
## How to Use
1. Clone this repository using the web URL.
```bash
git clone https://github.com/hyeonsieun/MathSpeech.git
```
2. To build the environment, run the following code
```bash
pip install -r requirements.txt
```
3. Place [the audio dataset and the transcription Excel file](https://drive.google.com/drive/folders/1M8_IVcesO2EwNcl9zwxY6UgqAmSODzgq?usp=sharing) inside the ASR folder.
4. Run the following code.
```bash
python ASR.py
```
5. Go to the Experiments folder
6. Move the 'MathSpeech_checkpoint.pth' from the following [link](https://drive.google.com/file/d/1m0cCpDDkOb7FltjLPVlg4ZCZSSSWZgS2/view?usp=sharing) into the Experiments folder.
7. Run the following code.
```bash
python MathSpeech_eval.py
```
8. If you want to run LLMs like GPT-4o or Gemini, you'll need to configure the environment settings such as the API key and endpoint.
9. You can also run the Ablation Study code from the Ablation_Study folder.

**Notes:** Here, example code for performing ASR using whisper-base and whisper-small is provided. If you want to use a different ASR model, you can modify that part of the code to use our MathSpeech.
Loading