<div align="center">
# 🧪 DECIMER Image Transformer 🖼️
### Deep Learning for Chemical Image Recognition using Efficient-Net V2 + Transformer
<p align="center">
<img src="https://github.com/Kohulan/DECIMER-Image_Transformer/blob/master/DECIMER_V2.png?raw=true" alt="DECIMER Logo" width="600">
</p>
[](https://opensource.org/licenses/MIT)
[](https://GitHub.com/Kohulan/DECIMER-Image_Transformer/graphs/commit-activity)
[](https://GitHub.com/Kohulan/DECIMER-Image_Transformer/issues/)
[](https://GitHub.com/Kohulan/DECIMER-Image_Transformer/graphs/contributors/)
[](https://www.tensorflow.org)
[](https://zenodo.org/badge/latestdoi/293572361)
[](https://decimer-image-transformer.readthedocs.io/en/latest/?badge=latest)
[](https://GitHub.com/Kohulan/DECIMER-Image_Transformer/releases/)
[](https://pypi.python.org/pypi/decimer/)
</div>
---
## 📚 Table of Contents
- [Abstract](#-abstract)
- [Method and Model Changes](#-method-and-model-changes)
- [Installation](#-installation)
- [Usage](#-usage)
- [Hand-drawn Model](#-decimer---hand-drawn-model)
- [Citation](#-citation)
- [Acknowledgements](#-acknowledgements)
- [Author](#-author-kohulan)
- [Project Website](#-project-website)
- [Research Group](#-research-group)
---
## 🔬 Abstract
<div style="background-color: #f0f0f0; padding: 15px; border-radius: 10px;">
The DECIMER 2.2 project tackles the OCSR (Optical Chemical Structure Recognition) challenge using cutting-edge computational intelligence methods. Our goal? To provide an automated, open-source software solution for chemical image recognition.
We've supercharged DECIMER with Google's TPU (Tensor Processing Unit) to handle datasets of over 1 million images with lightning speed!
</div>
---
## 🧠 Method and Model Changes
<table>
<tr>
<td width="50%">
<h3>🖼️ Image Feature Extraction</h3>
<p>Now utilizing EfficientNet-V2 for superior image analysis</p>
</td>
<td width="50%">
<h3>🔮 SMILES Prediction</h3>
<p>Employing a state-of-the-art transformer model</p>
</td>
</tr>
</table>
### 🚀 Training Enhancements
1. **TFRecord Files**: Lightning-fast data reading
2. **Google Cloud Buckets**: Efficient cloud storage solution
3. **TensorFlow Data Pipeline**: Optimized data loading
4. **TPU Strategy**: Harnessing the power of Google's TPUs
---
## 💻 Installation
```bash
# Create a conda wonderland
conda create --name DECIMER python=3.10.0 -y
conda activate DECIMER
# Equip yourself with DECIMER
pip install decimer
```
---
## 🎮 Usage
```python
from DECIMER import predict_SMILES
# Unleash the power of DECIMER
image_path = "path/to/your/chemical/masterpiece.jpg"
SMILES = predict_SMILES(image_path)
print(f"🎉 Decoded SMILES: {SMILES}")
```
---
## ✍️ DECIMER - Hand-drawn Model
<div style="background-color: #e6f7ff; padding: 15px; border-radius: 10px;">
🌟 **New Feature Alert!** 🌟
Our latest model brings the magic of AI to hand-drawn chemical structures!
[](https://doi.org/10.5281/zenodo.10781330)
</div>
---
## 📜 Citation
<div style="background-color: #f9f9f9; padding: 15px; border-radius: 10px;">
If DECIMER helps your research, please cite:
1. Rajan K, et al. "DECIMER.ai - An open platform for automated optical chemical structure identification, segmentation and recognition in scientific publications." *Nat. Commun.* 14, 5045 (2023).
2. Rajan, K., et al. "DECIMER 1.0: deep learning for chemical image recognition using transformers." *J Cheminform* 13, 61 (2021).
3. Rajan, K., et al. "Advancements in hand-drawn chemical structure recognition through an enhanced DECIMER architecture," *J Cheminform* 16, 78 (2024).
</div>
---
## 🙏 Acknowledgements
- A big thank you to [Charles Tapley Hoyt](https://github.com/cthoyt) for his invaluable contributions!
- Powered by Google's TPU Research Cloud (TRC)
<p align="center">
<img src="https://user-images.githubusercontent.com/30716951/220350828-913e6645-6a0a-403c-bcb8-160d061d4606.png" width="300">
</p>
---
## 👨🔬 Author: [Kohulan](https://kohulanr.com)
<p align="center">
<img src="https://github.com/Kohulan/DECIMER-Image-to-SMILES/raw/master/assets/DECIMER.gif" width="300">
</p>
---
## 🌐 Project Website
Experience DECIMER in action at [decimer.ai](https://decimer.ai), brilliantly implemented by [Otto Brinkhaus](https://github.com/OBrink)!
---
## 🏫 Research Group
<p align="center">
<a href="https://cheminf.uni-jena.de">
<img src="https://github.com/Kohulan/DECIMER-Image-to-SMILES/blob/master/assets/CheminfGit.png" width="300">
</a>
</p>
---
<div align="center">
### 📊 Project Analytics
</div>
Raw data
{
"_id": null,
"home_page": "https://github.com/Kohulan/DECIMER-Image_Transformer",
"name": "decimer",
"maintainer": "Kohulan Rajan, Otto Brinkhaus",
"docs_url": null,
"requires_python": ">=3.5",
"maintainer_email": "kohulan.rajan@uni-jena.de, otto.brinkhaus@uni-jena.de",
"keywords": null,
"author": "Kohulan Rajan",
"author_email": "kohulan.rajan@uni-jena.de",
"download_url": "https://files.pythonhosted.org/packages/08/55/3e0571ed9b738a6ebeb9c99dc316ea098643687209098a85407c37e1c83c/decimer-2.7.1.tar.gz",
"platform": null,
"description": "<div align=\"center\">\n\n# \ud83e\uddea DECIMER Image Transformer \ud83d\uddbc\ufe0f\n\n### Deep Learning for Chemical Image Recognition using Efficient-Net V2 + Transformer\n\n<p align=\"center\">\n <img src=\"https://github.com/Kohulan/DECIMER-Image_Transformer/blob/master/DECIMER_V2.png?raw=true\" alt=\"DECIMER Logo\" width=\"600\">\n</p>\n\n[](https://opensource.org/licenses/MIT)\n[](https://GitHub.com/Kohulan/DECIMER-Image_Transformer/graphs/commit-activity)\n[](https://GitHub.com/Kohulan/DECIMER-Image_Transformer/issues/)\n[](https://GitHub.com/Kohulan/DECIMER-Image_Transformer/graphs/contributors/)\n[](https://www.tensorflow.org)\n[](https://zenodo.org/badge/latestdoi/293572361)\n[](https://decimer-image-transformer.readthedocs.io/en/latest/?badge=latest)\n[](https://GitHub.com/Kohulan/DECIMER-Image_Transformer/releases/)\n[](https://pypi.python.org/pypi/decimer/)\n\n</div>\n\n---\n\n## \ud83d\udcda Table of Contents\n\n- [Abstract](#-abstract)\n- [Method and Model Changes](#-method-and-model-changes)\n- [Installation](#-installation)\n- [Usage](#-usage)\n- [Hand-drawn Model](#-decimer---hand-drawn-model)\n- [Citation](#-citation)\n- [Acknowledgements](#-acknowledgements)\n- [Author](#-author-kohulan)\n- [Project Website](#-project-website)\n- [Research Group](#-research-group)\n\n---\n\n## \ud83d\udd2c Abstract\n\n<div style=\"background-color: #f0f0f0; padding: 15px; border-radius: 10px;\">\n\nThe DECIMER 2.2 project tackles the OCSR (Optical Chemical Structure Recognition) challenge using cutting-edge computational intelligence methods. Our goal? To provide an automated, open-source software solution for chemical image recognition.\n\nWe've supercharged DECIMER with Google's TPU (Tensor Processing Unit) to handle datasets of over 1 million images with lightning speed!\n\n</div>\n\n---\n\n## \ud83e\udde0 Method and Model Changes\n\n<table>\n <tr>\n <td width=\"50%\">\n <h3>\ud83d\uddbc\ufe0f Image Feature Extraction</h3>\n <p>Now utilizing EfficientNet-V2 for superior image analysis</p>\n </td>\n <td width=\"50%\">\n <h3>\ud83d\udd2e SMILES Prediction</h3>\n <p>Employing a state-of-the-art transformer model</p>\n </td>\n </tr>\n</table>\n\n### \ud83d\ude80 Training Enhancements\n\n1. **TFRecord Files**: Lightning-fast data reading\n2. **Google Cloud Buckets**: Efficient cloud storage solution\n3. **TensorFlow Data Pipeline**: Optimized data loading\n4. **TPU Strategy**: Harnessing the power of Google's TPUs\n\n---\n\n## \ud83d\udcbb Installation\n\n```bash\n# Create a conda wonderland\nconda create --name DECIMER python=3.10.0 -y\nconda activate DECIMER\n\n# Equip yourself with DECIMER\npip install decimer\n```\n\n---\n\n## \ud83c\udfae Usage\n\n```python\nfrom DECIMER import predict_SMILES\n\n# Unleash the power of DECIMER\nimage_path = \"path/to/your/chemical/masterpiece.jpg\"\nSMILES = predict_SMILES(image_path)\nprint(f\"\ud83c\udf89 Decoded SMILES: {SMILES}\")\n```\n\n---\n\n## \u270d\ufe0f DECIMER - Hand-drawn Model\n\n<div style=\"background-color: #e6f7ff; padding: 15px; border-radius: 10px;\">\n\n\ud83c\udf1f **New Feature Alert!** \ud83c\udf1f\n\nOur latest model brings the magic of AI to hand-drawn chemical structures!\n\n[](https://doi.org/10.5281/zenodo.10781330)\n\n</div>\n\n---\n\n## \ud83d\udcdc Citation\n\n<div style=\"background-color: #f9f9f9; padding: 15px; border-radius: 10px;\">\n\nIf DECIMER helps your research, please cite:\n\n1. Rajan K, et al. \"DECIMER.ai - An open platform for automated optical chemical structure identification, segmentation and recognition in scientific publications.\" *Nat. Commun.* 14, 5045 (2023).\n2. Rajan, K., et al. \"DECIMER 1.0: deep learning for chemical image recognition using transformers.\" *J Cheminform* 13, 61 (2021).\n3. Rajan, K., et al. \"Advancements in hand-drawn chemical structure recognition through an enhanced DECIMER architecture,\" *J Cheminform* 16, 78 (2024).\n\n</div>\n\n---\n\n## \ud83d\ude4f Acknowledgements\n\n- A big thank you to [Charles Tapley Hoyt](https://github.com/cthoyt) for his invaluable contributions!\n- Powered by Google's TPU Research Cloud (TRC)\n\n<p align=\"center\">\n <img src=\"https://user-images.githubusercontent.com/30716951/220350828-913e6645-6a0a-403c-bcb8-160d061d4606.png\" width=\"300\">\n</p>\n\n---\n\n## \ud83d\udc68\u200d\ud83d\udd2c Author: [Kohulan](https://kohulanr.com)\n\n<p align=\"center\">\n <img src=\"https://github.com/Kohulan/DECIMER-Image-to-SMILES/raw/master/assets/DECIMER.gif\" width=\"300\">\n</p>\n\n---\n\n## \ud83c\udf10 Project Website\n\nExperience DECIMER in action at [decimer.ai](https://decimer.ai), brilliantly implemented by [Otto Brinkhaus](https://github.com/OBrink)!\n\n---\n\n## \ud83c\udfeb Research Group\n\n<p align=\"center\">\n <a href=\"https://cheminf.uni-jena.de\">\n <img src=\"https://github.com/Kohulan/DECIMER-Image-to-SMILES/blob/master/assets/CheminfGit.png\" width=\"300\">\n </a>\n</p>\n\n---\n\n<div align=\"center\">\n\n### \ud83d\udcca Project Analytics\n\n![Repobeats](https://repobeats.axiom.co/api/embed/bf532b7ac0d34137bdea8fbb82986828f86de065.svg \"Repobeats analytics image\")\n\n</div>\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "DECIMER 2.6.0: Deep Learning for Chemical Image Recognition using Efficient-Net V2 + Transformer",
"version": "2.7.1",
"project_urls": {
"Homepage": "https://github.com/Kohulan/DECIMER-Image_Transformer"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "e53244e2413fee2f8d0740d5bcb21dd3366378abe81b2f1189090f93b3a4a1fa",
"md5": "da3c6b1ae579e97003e4b23bfe5f79bb",
"sha256": "3ec3b1d972f71b9384048c748541c2d0e2726eb3c61d86d09473fb908d431908"
},
"downloads": -1,
"filename": "decimer-2.7.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "da3c6b1ae579e97003e4b23bfe5f79bb",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.5",
"size": 47575,
"upload_time": "2024-11-04T08:30:46",
"upload_time_iso_8601": "2024-11-04T08:30:46.695023Z",
"url": "https://files.pythonhosted.org/packages/e5/32/44e2413fee2f8d0740d5bcb21dd3366378abe81b2f1189090f93b3a4a1fa/decimer-2.7.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "08553e0571ed9b738a6ebeb9c99dc316ea098643687209098a85407c37e1c83c",
"md5": "cd454bcf6e6e2b473f9243a2be680d6e",
"sha256": "f52bb8e5861c401a5028d63b712c16db7b053f43163b4507b877635c3e85ead1"
},
"downloads": -1,
"filename": "decimer-2.7.1.tar.gz",
"has_sig": false,
"md5_digest": "cd454bcf6e6e2b473f9243a2be680d6e",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.5",
"size": 38564,
"upload_time": "2024-11-04T08:30:47",
"upload_time_iso_8601": "2024-11-04T08:30:47.785054Z",
"url": "https://files.pythonhosted.org/packages/08/55/3e0571ed9b738a6ebeb9c99dc316ea098643687209098a85407c37e1c83c/decimer-2.7.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-11-04 08:30:47",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "Kohulan",
"github_project": "DECIMER-Image_Transformer",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"requirements": [
{
"name": "efficientnet",
"specs": []
},
{
"name": "opencv-python",
"specs": []
},
{
"name": "pillow-heif",
"specs": []
},
{
"name": "pre-commit",
"specs": []
},
{
"name": "pystow",
"specs": []
},
{
"name": "tensorflow",
"specs": [
[
"<=",
"2.15.0"
],
[
">=",
"2.8.0"
]
]
}
],
"tox": true,
"lcname": "decimer"
}
JSON
