project setup

[mishradev1]

This PR establishes the foundational project structure and core utilities for the Intelligent Closed Caption Generation tool. It sets up a clean, modular architecture designed to support the three main goals - Sound Event Detection, Speaker Reaction Detection, and the CC Decision Engine.

Fixes #2

Key Changes

• Project Setup: Established a modular src/ directory structure with separate packages for utils/, detectors/, models/, engine/, and output/ - each mapped to a specific pipeline stage.

• Configuration System (config/settings.py): Created a centralized configuration module with documented defaults for all pipeline stages - audio extraction, sound event detection (YAMNet), reaction detection (MediaPipe), CC decision thresholds, and output formatting. This ensures all future modules share consistent, tunable parameters.

• Audio Extraction (src/utils/audio_extractor.py): Built an AudioExtractor class that uses FFmpeg directly via subprocess to strip audio from video files and convert to 16kHz mono WAV - the format required by YAMNet for sound event classification. Includes input validation for 8 video formats, proper error handling, and timeout protection.

• Comprehensive README: Added full project documentation with an architecture diagram, installation guide, usage examples, project structure breakdown, and tech stack overview.

• Dependencies & Packaging: Added requirements.txt with all planned dependencies (TensorFlow, OpenCV, MediaPipe, librosa, etc.) and setup.py with a cc-suggest console script entry point.

• Test Suite: Added 15 unit tests covering AudioExtractor initialization, input validation, extraction success/failure, timeouts, and output path generation. All tests use mocked FFmpeg - no external tools needed to run them.

How to Test

1. Pull this branch and create a virtual environment

2. Install dependencies:
   pip install -r requirements.txt

3. Run the test suite:
   python -m pytest tests/ -v

Expected: 15 passed

[mishradev1]

@abinash-sketch @keerthiseelan-planetread
Could you please review this initial setup and let me know if the project structure and direction align with the goals? The next PR I am working on will build on this to add the Sound Event Detection module (Goal 1) and Speaker Reaction Detection module (Goal 2).

[abinash-sketch]

let me know when can we connect.