Claude
0e73c78d7a
Simplification to focus on core alignment functionality: Removed: - WhisperXTranscribeNode class (143 lines) - All transcription-related functionality - Transcription workflow examples from README - Transcription node documentation Rationale: - Alignment is the unique value proposition - Users can use any transcription tool (Whisper, Faster-Whisper, etc.) - Reduces maintenance burden and code complexity - Clearer product focus: text-to-audio alignment only Updated: - Package description: now "text-audio alignment" instead of "transcription and alignment" - Features list: emphasizes alignment capabilities - README workflows: removed transcription examples - example_workflow.json: simplified to show plain text alignment Node Registry: - NODE_CLASS_MAPPINGS: only WhisperXAlignmentNode - NODE_DISPLAY_NAME_MAPPINGS: only alignment node The WhisperX Alignment node remains fully functional: - Plain text and JSON input support - Automatic text segmentation - Multi-language support - Word and character-level alignment - GPU acceleration Users can now: 1. Transcribe with their preferred tool 2. Use this node to get precise word-level timestamps 3. Leverage automatic text segmentation for plain transcripts Generated with Claude Code Co-Authored-By: Claude <noreply@anthropic.com>
60 lines
1.6 KiB
JSON
60 lines
1.6 KiB
JSON
{
|
|
"workflow_name": "WhisperX Text-Audio Alignment",
|
|
"description": "Example workflow showing how to align text transcripts with audio using WhisperX Alignment node",
|
|
"nodes": [
|
|
{
|
|
"node_type": "Load Audio (ComfyUI Official)",
|
|
"id": 1,
|
|
"inputs": {
|
|
"audio_path": "/path/to/your/audio.wav"
|
|
},
|
|
"outputs": {
|
|
"audio": "Connect to WhisperX Alignment node"
|
|
},
|
|
"notes": "Use ComfyUI's official Load Audio node (e.g., from ComfyUI-Advanced-Audio)"
|
|
},
|
|
{
|
|
"node_type": "WhisperX Alignment",
|
|
"id": 2,
|
|
"inputs": {
|
|
"audio": "Connect from Load Audio node",
|
|
"input_type": "plain_text",
|
|
"text_input": "Paste your transcript here or use JSON format",
|
|
"language": "en",
|
|
"auto_segment": true,
|
|
"max_chars_per_segment": 200,
|
|
"return_char_alignments": false,
|
|
"device": "auto"
|
|
},
|
|
"outputs": {
|
|
"aligned_segments": "Segments with accurate timestamps",
|
|
"word_segments": "Word-level timestamps",
|
|
"alignment_info": "Alignment statistics"
|
|
},
|
|
"notes": "For JSON input, set input_type to 'json' and provide segments from a transcription tool"
|
|
}
|
|
],
|
|
"connections": [
|
|
{
|
|
"from": "Load Audio (1).audio",
|
|
"to": "WhisperX Alignment (2).audio"
|
|
}
|
|
],
|
|
"example_output": {
|
|
"word_segments": [
|
|
{
|
|
"word": "Hello",
|
|
"start": 0.52,
|
|
"end": 0.89,
|
|
"score": 0.95
|
|
},
|
|
{
|
|
"word": "world",
|
|
"start": 1.05,
|
|
"end": 1.67,
|
|
"score": 0.97
|
|
}
|
|
]
|
|
}
|
|
}
|