Spaces:
Sleeping
Sleeping
Commit
Β·
a65e49e
1
Parent(s):
a1f56e0
update README.md
Browse files
README.md
CHANGED
|
@@ -16,6 +16,9 @@ pinned: false # Whether this Space should be pinned on your profile
|
|
| 16 |
## Project Overview
|
| 17 |
This is a **web-based Optical Character Recognition (OCR) application** built using Streamlit. The app supports both English and Hindi languages, allowing users to upload images and extract text using advanced OCR models.
|
| 18 |
|
|
|
|
|
|
|
|
|
|
| 19 |
## How the Application Works
|
| 20 |
1. Choose Language: Select either English or Hindi using the sidebar instructions.
|
| 21 |
2. Upload Image: Use the file uploader to input an image in JPG, PNG, or JPEG format.
|
|
@@ -49,7 +52,7 @@ This is a **web-based Optical Character Recognition (OCR) application** built us
|
|
| 49 |
|
| 50 |
# Description
|
| 51 |
|
| 52 |
-
This web application
|
| 53 |
|
| 54 |
# GOT OCR 2.0 Model Overview
|
| 55 |
|
|
@@ -62,15 +65,29 @@ Note: Currently, the model does not support all languages. Fine-tuning is requir
|
|
| 62 |
|
| 63 |
For more technical details about the model architecture and usage, visit the [GOT OCR 2.0 Model Documentation](https://github.com/Ucas-HaoranWei/GOT-OCR2.0/?tab=readme-ov-file#general-ocr-theory-towards-ocr-20-via-a-unified-end-to-end-model).
|
| 64 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
|
| 66 |
## Deployment
|
| 67 |
-
To deploy the application to a cloud platform(Hugging Face)
|
| 68 |
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
.
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
|
| 76 |
|
|
|
|
| 16 |
## Project Overview
|
| 17 |
This is a **web-based Optical Character Recognition (OCR) application** built using Streamlit. The app supports both English and Hindi languages, allowing users to upload images and extract text using advanced OCR models.
|
| 18 |
|
| 19 |
+
## Live Demo
|
| 20 |
+
You can access the live demo of the application at: [https://huggingface.co/spaces/Trisandhya/GOT-OCR-WEB-APP](https://huggingface.co/spaces/Trisandhya/GOT-OCR-WEB-APP)
|
| 21 |
+
|
| 22 |
## How the Application Works
|
| 23 |
1. Choose Language: Select either English or Hindi using the sidebar instructions.
|
| 24 |
2. Upload Image: Use the file uploader to input an image in JPG, PNG, or JPEG format.
|
|
|
|
| 52 |
|
| 53 |
# Description
|
| 54 |
|
| 55 |
+
This web application facilitates the conversion of images to text using the GOT OCR 2.0 Model for English text extraction. While the GOT model excels in processing English content, fine-tuning it on a Hindi dataset is not feasible on a CPU. Therefore, for Hindi text extraction, we utilize EasyOCR, which provides effective performance for this language.
|
| 56 |
|
| 57 |
# GOT OCR 2.0 Model Overview
|
| 58 |
|
|
|
|
| 65 |
|
| 66 |
For more technical details about the model architecture and usage, visit the [GOT OCR 2.0 Model Documentation](https://github.com/Ucas-HaoranWei/GOT-OCR2.0/?tab=readme-ov-file#general-ocr-theory-towards-ocr-20-via-a-unified-end-to-end-model).
|
| 67 |
|
| 68 |
+
## Folder Structure
|
| 69 |
+
|
| 70 |
+
<pre>
|
| 71 |
+
.
|
| 72 |
+
βββ app.py
|
| 73 |
+
βββ requirements.txt
|
| 74 |
+
βββ README.md
|
| 75 |
+
|
| 76 |
+
</pre>
|
| 77 |
|
| 78 |
## Deployment
|
|
|
|
| 79 |
|
| 80 |
+
# To deploy the application to a Hugging Face cloud platform
|
| 81 |
+
|
| 82 |
+
1. Use GitHub Actions: Set up GitHub Actions in your repository to automate the deployment process to Hugging Face Spaces.
|
| 83 |
+
|
| 84 |
+
2. Follow Documentation: For detailed instructions on setting up GitHub Actions for Hugging Face Spaces, refer to the [Hugging Face Spaces GitHub Actions Documentation](https://huggingface.co/docs/hub/spaces-github-actions).
|
| 85 |
+
|
| 86 |
+
|
| 87 |
+
## Contact
|
| 88 |
+
For any further queries or assistance, feel free to reach out via email at:[devitrisandhya@gmail.com](devitrisandhya@gmail.com)
|
| 89 |
+
|
| 90 |
+
|
| 91 |
+
|
| 92 |
|
| 93 |
|