Zero Shot Audio Classifier

Type: huggingface.audio_classification.ZeroShotAudioClassifier

Namespace: huggingface.audio_classification

Description

Classifies audio into categories without the need for training data. audio, classification, labeling, categorization, zero-shot

Use cases:
- Quickly categorize audio without training data
- Identify sounds or music genres without predefined labels
- Automate audio tagging for large datasets

Properties

Property Type Description Default
model hf.zero_shot_audio_classification The model ID to use for the classification {'type': 'hf.zero_shot_audio_classification', 'repo_id': '', 'path': None, 'variant': None, 'allow_patterns': None, 'ignore_patterns': None}
audio audio The input audio to classify {'type': 'audio', 'uri': '', 'asset_id': None, 'data': None}
candidate_labels str The candidate labels to classify the audio against, separated by commas ``

Outputs

Output Type Description
output Dict[str, float]  

Metadata

Browse other nodes in the huggingface.audio_classification namespace.