Update README.md
Browse files
README.md
CHANGED
|
@@ -1,98 +1,12 @@
|
|
| 1 |
-
Absolutely! Here’s the full rephrased content you can easily copy and paste:
|
| 2 |
-
|
| 3 |
-
⸻
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
# Image Description and Audio Transcript App
|
| 8 |
-
|
| 9 |
-
An AI-powered web app that identifies objects in images and converts the generated descriptions into speech using Hugging Face Transformers.
|
| 10 |
-
|
| 11 |
-
---
|
| 12 |
-
|
| 13 |
-
## Overview
|
| 14 |
-
This project showcases how to build a pipeline using:
|
| 15 |
-
- **BLIP** for image captioning
|
| 16 |
-
- **gTTS** (Google Text-to-Speech) for audio generation
|
| 17 |
-
- **Gradio** for the user interface and deployment on Hugging Face Spaces
|
| 18 |
-
|
| 19 |
-
---
|
| 20 |
-
|
| 21 |
-
## What It Does
|
| 22 |
-
- Upload an image → Get an AI-generated description
|
| 23 |
-
- Automatically convert the description into audio
|
| 24 |
-
- Built with accessibility in mind for users with visual impairments
|
| 25 |
-
- Runs on a clean, responsive web UI using **Gradio**
|
| 26 |
-
|
| 27 |
-
---
|
| 28 |
-
|
| 29 |
-
## Tech Stack
|
| 30 |
-
- **Language**: Python 3.7+
|
| 31 |
-
- **AI Models**:
|
| 32 |
-
- `Salesforce/blip-image-captioning-base` – for generating image captions
|
| 33 |
-
- `gtts` – for converting text into speech
|
| 34 |
-
- **Frameworks/Libraries**:
|
| 35 |
-
- `torch` – powering the models
|
| 36 |
-
- `transformers` – loading and running pre-trained models
|
| 37 |
-
- `gradio` – creating the interactive frontend
|
| 38 |
-
- `Pillow`, `matplotlib`, `inflect` – for image handling and fine-tuning the output
|
| 39 |
-
|
| 40 |
---
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
2. (Optional) Create a virtual environment:
|
| 50 |
-
|
| 51 |
-
python -m venv venv
|
| 52 |
-
source venv/bin/activate # For Windows: venv\Scripts\activate
|
| 53 |
-
|
| 54 |
-
3. Install dependencies:
|
| 55 |
-
|
| 56 |
-
pip install torch transformers gtts gradio Pillow matplotlib inflect
|
| 57 |
-
|
| 58 |
-
If you’re using Hugging Face Spaces, simply include a requirements.txt file with those packages.
|
| 59 |
-
|
| 60 |
-
⸻
|
| 61 |
-
|
| 62 |
-
How to Run
|
| 63 |
-
|
| 64 |
-
Locally:
|
| 65 |
-
|
| 66 |
-
python object_detection.py
|
| 67 |
-
|
| 68 |
-
Then visit: http://127.0.0.1:7860 in your browser.
|
| 69 |
-
|
| 70 |
-
On Hugging Face:
|
| 71 |
-
Just upload all files (including requirements.txt) to your Space. It’ll launch automatically.
|
| 72 |
-
|
| 73 |
-
⸻
|
| 74 |
-
|
| 75 |
-
Customizations
|
| 76 |
-
|
| 77 |
-
You can tweak parameters (like host, port, or debug settings) directly in the script if needed. For example:
|
| 78 |
-
|
| 79 |
-
gr.Interface(...).launch(server_name="0.0.0.0", server_port=7860, debug=True)
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
⸻
|
| 84 |
-
|
| 85 |
-
Credits
|
| 86 |
-
• Hugging Face for the BLIP model
|
| 87 |
-
• Google for gTTS
|
| 88 |
-
• Gradio for simplifying deployment and UI creation
|
| 89 |
-
|
| 90 |
-
⸻
|
| 91 |
-
|
| 92 |
-
License
|
| 93 |
-
|
| 94 |
-
MIT License – Feel free to use, share, and modify.
|
| 95 |
-
|
| 96 |
---
|
| 97 |
|
| 98 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
title: Object Detection App
|
| 3 |
+
emoji: 🧠
|
| 4 |
+
colorFrom: indigo
|
| 5 |
+
colorTo: blue
|
| 6 |
+
sdk: gradio
|
| 7 |
+
sdk_version: "4.20.0"
|
| 8 |
+
app_file: app.py
|
| 9 |
+
pinned: false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
---
|
| 11 |
|
| 12 |
+
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|