feat(transcription): add Whisper transcriber and audio utilities

- Add WhisperTranscriber wrapper for stable-ts/faster-whisper
- Add audio utilities for ffmpeg/ffprobe operations
- Add translator for two-stage translation workflow
- Support CPU/GPU with graceful degradation

backend/transcription/__init__.py

@@ -0,0 +1,5 @@
+"""Whisper transcription module."""
+from backend.transcription.transcriber import WhisperTranscriber
+from backend.transcription.translator import SRTTranslator, translate_srt_file
+
+__all__ = ["WhisperTranscriber", "SRTTranslator", "translate_srt_file"]