Papers
arxiv:2609.00055

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

Published on Aug 30
Authors:
,
,
,

Abstract

A framework aligns self-supervised respiratory encoders with medical terminology via synthetic reports and contrastive learning to enable high-performing zero-shot clinical diagnosis.

Self-supervised respiratory encoders lack semantic grounding in clinical domain needed for zero-shot inference, limiting their utility without task-specific labeled data. We propose a framework that aligns these encoders with medical terminology in a shared latent space turning them into a zero-shot-capable foundation model. To address paired data scarcity, we use a medical LLM to synthesize structured reports from metadata, creating dense semantic anchors for contrastive learning. Our training combines a sigmoid-based contrastive loss with encoder's native SSL objective and similarity-aware negative sampling to sharpen pathological boundaries. Across 9 tasks on 6 datasets, our method achieves a 61.3% mean zero-shot AUC, surpassing CLAP (51.4%) and Qwen2-Audio (54.9%) while reaching the highest linear probing AUC (71.6%) with only 43% of data used by full-scale baselines, showing that structured semantic alignment outperforms large-scale, general-purpose models in clinical diagnostics.

Community

Sign up or log in to comment

Get this paper in your agent:

hf papers read 2609.00055
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2609.00055 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2609.00055 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2609.00055 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.