🎥 AI-Powered Video Analysis System

🚀 AI-based Video Processing & Analysis System with Django, Celery, WebSockets, OpenAI GPT-4, Whisper, OCR, and YOLOv8 for video transcript generation, object detection, speech-to-text, and AI-powered chat.

📌 Features

✅ Upload & Process Videos – Extract audio, text, and objects

✅ Real-time Progress Updates – WebSockets for instant feedback

✅ AI Chat – Ask questions about the processed video

✅ Multi-user API Key Support – Users enter their own OpenAI API key

✅ Database Storage – Save transcripts & analysis reports

✅ WebSocket-powered Chat – Realtime AI Q&A

✅ Download Reports – Transcripts & analysis reports available

---

🏗️ Directory Structure

├── video_analysis_project/
│   ├── dump.rdb
│   ├── requirements.txt
│   ├── yolov8n.pt
│   ├── db.sqlite3
│   ├── README.md
│   ├── database.env
│   ├── docker-compose.yml
│   ├── manage.py
│   ├── celery_tasks.py
│   ├── app/
│   │   ├── video_processing.py
│   │   ├── tasks.py
│   │   ├── models.py
│   │   ├── database.py
│   │   ├── chat_processor.py
│   │   ├── consumers.py
│   │   ├── apps.py
│   │   ├── admin.py
│   │   ├── routing.py
│   │   ├── tests.py
│   │   ├── urls.py
│   │   └── views.py
│   ├── config/
│   │   ├── asgi.py
│   │   ├── celery.py
│   │   ├── settings.py
│   │   └── urls.py
│   └── Dockerfile/
│   ├── staticfiles/
│   │   ├── styles.css
│   │   └── script.js
│   │   ├── admin/
│   │   │   ├── css/
│   │   │   ├── js/
│   │   │   │   ├── admin/
│   │   │   │   ├── vendor/
│   │   │   ├── img/
│   │   │   │   ├── gis/
│   │   ├── rest_framework/
│   │   │   ├── css/
│   │   │   ├── js/
│   │   │   ├── docs/
│   │   │   │   ├── css/
│   │   │   │   ├── js/
│   │   │   │   ├── img/

│   │   │   ├── img/
│   │   │   ├── fonts/
│   ├── static/
│   │   ├── styles.css
│   │   └── script.js
│   ├── templates/
│   │   ├── index.html
│   │   └── reports.html
│   ├── media/
│   │   └── progress.json
│   │   ├── transcripts/
│   │   │   ├── Screen Recording 2025-02-08 at 9.49.34PM.mov_transcript.txt
│   │   │   └── Screen Recording 2025-02-13 at 2.04.42AM.mov_transcript.txt
│   │   ├── analysis/
│   │   │   ├── Screen Recording 2025-02-13 at 2.04.42AM.mov_analysis.txt
│   │   │   └── Screen Recording 2025-02-08 at 9.49.34PM.mov_analysis.txt
│   │   └── frames/
│   │   ├── uploads/
│   │   │   ├── Screen Recording 2025-02-08 at 9.49.34PM.mov
│   │   │   └── Screen Recording 2025-02-13 at 2.04.42AM.mov
│   │   └── chunks/
│   │   └── audio/

---

🛠️ Installation & Setup

1️⃣ Clone the Repository

git clone https://github.com/sachinarora/video_analysis_project.git
cd video_analysis_project

2️⃣ Setup Virtual Environment

python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts�ctivate

3️⃣ Install Dependencies

pip install -r requirements.txt

4️⃣ Setup PostgreSQL Database

• DB Name: postgres
• User: postgres
• Password: password

Update DATABASES in config/settings.py:

DATABASES = {
    'default': {
        'ENGINE': 'django.db.backends.postgresql',
        'NAME': 'postgres',
        'USER': 'postgres',
        'PASSWORD': 'password',
        'HOST': 'localhost',
        'PORT': '5432',
    }
}

5️⃣ Apply Migrations

python manage.py makemigrations
python manage.py migrate

6️⃣ Start Redis Server (For Celery)

redis-server

7️⃣ Start Celery Worker

celery -A config worker --loglevel=info

8️⃣ Run Django ASGI Server (For WebSockets)

daphne -b 0.0.0.0 -p 8000 config.asgi:application

---

📖 API Endpoints

Method	Endpoint	Description
POST	/upload_video/	Upload & process a new video
GET	/progress/	Get real-time processing progress
POST	/chat/	Ask AI questions about the video
GET	/reports/	View past analysis reports

---

🔧 Usage

1️⃣ Upload & Process a Video

1. Open http://127.0.0.1:8000/
2. Click Upload Video → Select a .mp4 file
3. Processing starts & progress is updated in real-time

2️⃣ View Live Progress

• WebSockets update UI dynamically
• Shows current processing step: "Extracting Audio", "Running OCR" etc.

3️⃣ Ask AI Questions

• Enter an OpenAI API Key
• Use chat box to ask "What objects are in this video?", etc.
• AI responds using transcript, OCR, & detected objects

4️⃣ Download Report

• Once processing is complete, download:
  • Transcript (media/transcripts/)
  • Analysis Report (media/analysis/)

---

📞 Contact & Support

📧 Email: sachnaror@gmail.com 🌍 GitHub: github.com/sachnaror 🌐 Website: https://about.me/sachin-arora

🚀 Happy Coding! 🎬💡

Commands:

First, stop existing processes, run:

pkill -f runserver
pkill -f redis
pkill -f celery

### and then..

redis-server &
python manage.py runserver &
celery -A config worker --loglevel=info &

1️⃣ Restart Redis

brew services restart redis

2️⃣ Apply migrations

python manage.py makemigrations python manage.py migrate

3️⃣ Start Celery Worker

celery -A config worker --loglevel=info

4️⃣ Restart Django ASGI Server

daphne -b 0.0.0.0 -p 8000 config.asgi:application