Commit History

  1. Add NaN/Inf safety check to prevent JSON serialization errors

    70820b4
  2. Fix torch version for PyPI compatibility

    69e411d
  3. Fix model loading: upgrade transformers to 5.0.0 and handle safetensors metadata issue

    f229a41
  4. Fix: Apply compatibility patches before speechbrain import

    e53bdb6
  5. Fix Dockerfile: Copy all files to include src and config

    d399729
  6. Prioritize root in sys.path and add debug info

    35456ff
  7. Fix Import Error: Add __init__.py files and debug logs

    d72a730
  8. Fix ModuleNotFoundError by adding root to sys.path

    08f0d4c
  9. Refactor project structure to Unified AI Voice Detection System

    b3f89f5
  10. Fix SpeechBrain compatibility with newer torchaudio by mocking list_audio_backends

    0b7673f
  11. Fix model loading error by updating torch version and improving logging

    6a16e92
  12. Add fastapi to requirements.txt

    4153c5e
  13. Heavy & Accurate: Integrated SpeechBrain VAD + MMS-300M pipeline

    62f98bb
  14. Upgrade: Switch to MMS-300M (XLS-R) for robust multilingual deepfake detection

    cb14a1d
  15. Fix: Require heuristic evidence for AI verdict on non-English audio to prevent false positives

    0d64f25
  16. Implement Conservative Consensus: SNR Analysis and Human Liveness Veto to eliminate false positives

    f95b090
  17. Implement supreme Hybrid AI detection: sliding window, spectral and temporal features, and language-aware confidence.

    04f3da2
  18. Add comprehensive debug logging to track AI classification confidence calculations

    9d2870b
  19. Upgrade to specialized deepfake detection model and optimize audio processing

    f07936b
  20. Fix: Align prediction label with pitch variance heuristic

    b1df80d
  21. Optimize: Truncate audio to 5s & add per-step timing logs

    28dad85
  22. Fix: Final audio processing fixes and enhanced logging

    a218549
  23. Fix: Add debug logging and robust padding to audio processing

    1f10fcd
  24. Fix: Robust Base64 decoding

    4b4453a
  25. Fix HF Spaces metadata

    b5ac8e5
  26. Configure Dockerfile for Hugging Face Spaces

    1f25e74
  27. Feat: ADD Dockerized the API

    5516b13
  28. Initial commit: Multilingual Voice Detector with XLS-R Support

    2a4d245