Papers/2609.00055
🧪 Test?View on arXiv

Zero-Shot Respiratory Sound Classification through LLM-Augmented Audio-Text Alignment

Not provided

zero-shot learningmedical AIcontrastive learningaudio classification
2609.00055
Builder Relevance
80%
1h ago

Abstract

The paper proposes a framework that enhances self-supervised respiratory encoders by aligning them with medical terminology, enabling zero-shot inference in clinical diagnostics.

Reality Card

Core Claim

The proposed framework achieves a 61.3% mean zero-shot AUC, outperforming existing models while using only 43% of the data required by full-scale baselines.

Method / Result

Achieved a 61.3% mean zero-shot AUC across 9 tasks on 6 datasets.

Limitations

The reliance on synthesized structured reports from metadata may limit reproducibility due to potential variability in report quality.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.
← Back to all papers