v1.0.0 — Official Release with Full CLI and Documentation

This release marks the first stable version of the VideoSemanticRepresentation framework.

🚀 Highlights

• Complete Command-Line Interface (CLI)

  • Added full argument parsing system with subcommands:
    • feature: extract and save semantic features from videos.
    • default mode: perform feature-based video retrieval.
  • Supports multiple input files, interactive overwrite prompts, and flexible output paths.

• Feature Extraction and Caching

  • GPU-accelerated 3D convolution–based semantic representation.
  • Automatic feature caching to speed up repeated runs.
  • Consistent save behavior even when cached data is reused.

• Documentation and Structure

  • Updated README.md to v1.0.0 with unified usage and examples.
  • Revised all module docstrings for style consistency.
  • Cleaned up codebase for maintainability and modularity.

📦 Build and Usage

# Extract video features
python -m src feature <input_video1> [<input_video2> ...] [-o <output_path>]

# Retrieve similar segment
python -m src <reference_video> -q <query_video1> [<query_video2> ...]