CaReAQA: A Cardiac and Respiratory Audio Question Answering Model for Open-Ended Diagnostic Reasoning

Aaqib Saeed; Lin-Lin Chen; Neil Zeghidour; Tsai-Ning Wang

REVIEW 4 major objections 5 minor 1 cited by

CaReAQA: A Cardiac and Respiratory Audio Question Answering Model for Open-Ended Diagnostic Reasoning

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read CaReAQA is an audio-language model that takes heart or lung sounds and a natural-language question and generates free-form diagnostic answers; the paper claims it outperforms general-purpose audio-language models on open-ended and…

desk verdict CaReSound is a useful new benchmark and the closed-ended results are credible, but the abstract's 86.2% accuracy claim mislabels BERTScore and the open-ended evaluation needs human calibration. read the letter →

arxiv 2505.01199 v2 pith:MLYPBCDR submitted 2025-05-02 cs.LG

Tsai-Ning Wang , Lin-Lin Chen , Neil Zeghidour , Aaqib Saeed This is my paper · ORCID

classification cs.LG

keywords medicalaudioquestionansweringdiagnosticreasoningheartandlungsoundsaudio-languagemodelCaReSoundbenchmarkopen-endedauscultationself-supervisedencoder

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

CaReAQA aims to make audio-based diagnostics open-ended: instead of classifying a recording into a fixed disease label, the model takes a heart or lung sound plus a clinician-style question and generates a free-form diagnostic answer. To train and test it, the paper assembles CaReSound, a benchmark of 16,273 recordings drawn from five public cardiac and respiratory datasets, with 34,792 question-answer pairs that an LLM generated from each recording's metadata. On the open-ended test split, the paper reports a combined BERTScore of 86.2 and a semantic accuracy of 70.6% measured by GPT-4o judgments, ahead of general-purpose audio-language baselines; the abstract's '86.2% accuracy' is the BERTScore, not the accuracy column. The same model reaches 56.9% average accuracy on four previously unseen closed-ended classification datasets. The broader claim is that aligning a self-supervised audio encoder with an LLM through synthetic QA data is a viable route to diagnostic reasoning from auscultation sounds.

What carries the argument

The mechanism is a three-part alignment stack. An OPERA contrastive audio encoder converts a 5-second log-mel spectrogram into embeddings; a transformer-based mapper projects those embeddings into the input space of LLaMA-3.2-3B; and the text question and audio embeddings are concatenated, positionally encoded, and fed through the LLM, which is fine-tuned with LoRA (rank 8) to maximize $p_\theta(A_i \mid X_q, X_a, A_{<i})$. The other load-bearing piece is CaReSound itself: GPT-4o writes three or more QA pairs per recording from metadata alone, so the dataset is large (34,792 pairs) but entirely synthetic.

What would settle it

Sample 200 open-ended test responses, have two clinicians independently mark each as semantically matching the reference or not, and compare their marks with GPT-4o's Yes/No verdicts: if clinician agreement with GPT-4o is low, or if GPT-4o says 'Yes' disproportionately to answers phrased like its own output, the reported 70.6% accuracy does not measure diagnostic agreement.

Watch

Extended reading notes

Core claim

The paper's central claim is that open-ended diagnostic reasoning over cardiac and respiratory audio can be learned end-to-end by a model that maps audio embeddings into an LLM's token space and fine-tunes the whole stack on synthetic question-answer pairs. The evidence is the CaReSound benchmark and the performance comparison: CaReAQA scores 86.2 BERTScore and 70.6% GPT-4o-judged accuracy on combined open-ended questions, versus at best 56.8 BERTScore and 24.5% accuracy among the compared general-purpose baselines; on unseen closed-ended tasks it averages 56.9% accuracy. The paper also claims the design choices matter: LoRA fine-tuning of the LLM, a transformer mapper, and fine-tuning the OPERA audio encoder each outperform their alternatives in ablations.

Load-bearing premise

The entire open-ended evaluation depends on trusting GPT-4o to have written the reference answers correctly and to judge correctly whether the model's free-form responses mean the same thing as those references.

Editorial extensions

If this is right

Open-ended auscultation question answering is trainable without manually written clinical QA pairs: an LLM can generate the training data from structured metadata.
Domain-specific audio-language alignment transfers beyond the training label set: the same model answers questions in-domain and classifies four unseen datasets at 56.9% average accuracy.
On this benchmark, fine-tuning the audio encoder and the LLM with LoRA matters more than the choice of mapper or a larger parametric adapter capacity.
Failure cases mean the model is not yet a diagnostic instrument: overlapping auscultation patterns, rare diseases, and noisy recordings still produce wrong answers.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Because the reference answers and the accuracy judgments both come from GPT-4o, the open-ended numbers measure agreement with one LLM's text style as much as diagnostic correctness; a clinician-labeled audit set would separate these.
The synthetic QA generation from metadata only means the model might answer from textual priors such as chest location or demographics rather than acoustic evidence; questions whose answers are deliberately contradicted by the audio would test this.
The recipe is portable: any medical or industrial sound domain with rich metadata could get a QA benchmark the same way, and the per-dataset variance suggests such benchmarks need per-domain reporting, not just aggregates.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Thanks for the pointer. Quick take: CaReSound is a genuinely useful new resource for the medical-audio QA subfield, and the model is a competent application of standard audio-LLM recipes. But the abstract's headline number is wrong — 86.2 is BERTScore, not accuracy — and the open-ended evaluation leans on GPT-4o as both answer generator and judge, so the open-ended accuracy numbers are not established as diagnostic accuracy. That doesn't sink the paper; the closed-ended generalization results rely on real labels and suggest real learning. It needs fixing before I'd trust the reasoning claims.

What's actually new: CaReSound, assembled from five public datasets (ICBHI, KAUH, CirCor, SPRSound, ZCHSound) with metadata and 34,792 GPT-4o-generated QA pairs. The data and model are released, which is a concrete contribution. The modeling itself — OPERA encoder, transformer mapper, LoRA-tuned LLaMA-3.2-3B — is not architecturally novel, but the medical-audio application is new and the ablation study (PEFT methods, mapping types, encoder choices) is more thorough than many papers in this space. The closed-ended results on four unseen datasets (56.9% average accuracy, beating baselines) are the most convincing part; those use human labels and suggest the audio encoder is capturing clinically meaningful structure.

Where it gets soft. The abstract says “86.2% accuracy on open-ended diagnostic reasoning tasks.” Table 3 shows the combined Accuracy is 70.6; 86.2 is the BERTScore. That is a direct mislabel, not a rounding difference. The accuracy metric itself is GPT-4o answering Yes/No on whether a prediction matches a ground-truth answer that GPT-4o also generated from metadata. LLM-as-judge is common, but without a human-calibration study — agreement on a sample, error bars, or at least a qualitative error analysis — the open-ended numbers are hard to interpret, especially for medical claims. Dataset counts are also inconsistent: Section 3.5 says 12,673 samples while Table 1 sums to 16,273. Minor but sloppy. The failure-case appendix is a point in the paper's favor, but those failures highlight exactly what should be checked against human judgment: is “holosystolic high-pitched harsh” meaningfully different from “holosystolic low-pitched blowing”, and would a clinician or even the judge call it correct?

Bottom line: this is a solid benchmark paper with a reasonable model and concrete, fixable evaluation problems. The benchmark alone may be worth the paper, and the closed-ended results support the model's basic competence. The headline needs correction and the open-ended evaluation needs human grounding. I would send it to peer review — the benchmark deserves scrutiny and the community would benefit from a tightened evaluation. I'd bring it to reading group mainly to discuss evaluation design.

Referee Report

4 major / 5 minor

Summary. The paper introduces CaReAQA, an audio-language model for cardiac and respiratory audio question answering, and CaReSound, a benchmark dataset assembled from five public medical audio collections. QA pairs for training and evaluation are generated by GPT-4o from textual metadata, and the model combines a fine-tuned OPERA audio encoder, a transformer mapper, and LLaMA-3.2-3B with LoRA. The authors report open-ended QA results using BERTScore, METEOR, and a GPT-4o-based semantic 'Accuracy' metric, plus closed-ended classification accuracy on four unseen datasets. They claim CaReAQA outperforms existing audio-language baselines on open-ended diagnostic reasoning and generalizes to closed-ended classification.

Significance. If substantiated, the paper would provide a useful open resource: the CaReSound dataset and model are released on Hugging Face, and the ablation studies (LoRA rank, PEFT methods, mapping types, audio encoders) are systematic and informative. The closed-ended generalization results use real labels and provide credible evidence that the model captures clinically relevant audio categories. The explicit failure-case analysis in Appendix F and the statement in the conclusion that the model is not intended for direct clinical deployment are appropriate. The main significance hinges on the open-ended diagnostic reasoning claim, and that is exactly where the current evaluation is not yet convincing.

major comments (4)

[Abstract and Section 4.2, Table 3] The abstract's claim of '86.2% accuracy on open-ended diagnostic reasoning tasks' is not supported by Table 3: the All* Accuracy column is 70.6, while 86.2 is the All* BERTScore. This is a load-bearing mislabel of the headline result and must be corrected in the abstract, conclusion, and any summary of results; the text in Section 4.2 already correctly refers to 'an average BERTScore of 86.2', so the discrepancy appears to be a reporting error rather than a substantive result, but it must be fixed.
[Section 3.3.1, Appendix B, Appendix C] The open-ended Accuracy metric is computed by prompting GPT-4o to judge whether a prediction is 'same/similar' to a reference (Appendix B), but the references themselves were generated by GPT-4o from metadata during dataset construction (Section 3.2, Appendix C), and CaReAQA was fine-tuned on those same GPT-4o-generated answers. This creates a closed loop: CaReAQA's predictions are in-distribution for the judge, while baseline models produce out-of-distribution phrasings. The paper reports no human evaluation, no inter-annotator agreement, and no confidence intervals. As a result, the reported 70.6% accuracy and 86.2 BERTScore do not yet establish diagnostic correctness for open-ended reasoning. Please add independent human or clinical validation on a sample, report judge agreement statistics, and temper the conclusions accordingly.
[Section 3.5 vs Table 1] The dataset size is inconsistent. Section 3.5 states that the dataset contains 12,673 samples and is split into 10,138 training and 2,535 test samples, but Table 1 reports 16,273 samples in the All column, which matches the sum of the per-dataset rows (6,899 + 337 + 5,282 + 1,259 + 2,496). Please clarify which total is correct and re-state the split consistently. In addition, Table 2 reports BMD as having 5 classes while Section 3.3.2 describes the BMD-HS dataset as having six categories; these must be reconciled.
[Section 4.2, Table 3] The comparison with general-purpose audio-language models is not fully controlled. CaReAQA is fine-tuned on CaReSound, whereas LTU, Pengi, Qwen2-Audio, and GAMA appear to be evaluated without any fine-tuning on the training distribution; the LLM-without-audio and Cascaded baselines are not trained under the same protocol either. The claim that CaReAQA 'outperforms baseline models' should be qualified as 'outperforms zero-shot baselines' unless the baselines are fine-tuned under an equivalent protocol, or an explicit statement is added confirming the absence of fine-tuning for those baselines.

minor comments (5)

[Section 4.1] The citation for AugLy appears as an unresolved '(?)' in the text; please add the full reference.
[Section 4.4] The heading 'Effect of LoRA Rank.' is duplicated in the text immediately before Table 7; please remove the duplicate.
[Throughout] The dataset name is written inconsistently as both 'CirCor' and 'CIRCOR', and the model name appears as both 'Gama' and 'GAMA'; please use consistent spellings.
[Section 3.3.1] The metric called 'Accuracy' is a GPT-4o binary judgment of semantic equivalence, not correctness in the usual classification sense; please rename it (e.g., 'semantic acceptance rate') or clearly define it in every occurrence to avoid confusion with the classification accuracy used in Table 4.
[Figure 1 caption] The caption has a spacing typo: 'Figure 1:Spectrograms' should be 'Figure 1: Spectrograms'.

Circularity Check

1 steps flagged · score 4.0 of 10

Open-ended accuracy is a GPT-4o self-agreement measure, not an independent diagnostic metric; the headline 86.2% is BERTScore, while the GPT-4o-judged Accuracy is 70.6%.

self definitional [Sec. 3.2, Sec. 3.3.1, Appendix B, Appendix C; Abstract vs Table 3]
"We employ GPT-4o (OpenAI, 2024) to generate question-answer (QA) pairs based on metadata and annotations extracted from the datasets. ... we compute an Accuracy score by comparing the model’s predictions against the ground truth using GPT-4o (OpenAI, 2024). We prompt the GPT-4o model via API to compare whether the prediction and ground truth are the same, providing a binary judgment “Yes” or “No” for each pair."

The open-ended Accuracy is defined as the rate at which GPT-4o answers “Yes” when asked whether the model’s prediction is “same/similar” to the reference. The references are themselves GPT-4o generations from metadata (Appendix C), and CaReAQA is fine-tuned to maximize the likelihood of those GPT-4o-generated answers. Thus the metric measures agreement with the same system that produced both the training targets and the evaluation references; a model that imitates GPT-4o’s phrasing is rewarded regardless of clinical correctness. The headline “86.2% accuracy” is also the BERTScore column of Table 3 (All*), while the GPT-4o-judged Accuracy column for All* is 70.6, so the number quoted as accuracy is not even the Accuracy column.

full rationale

The only substantive circularity concern is the self-referential open-ended evaluation: GPT-4o creates the reference answers from metadata, CaReAQA is trained on those answers, and GPT-4o then judges whether predictions are “same/similar” to those references. This makes the reported open-ended Accuracy a measure of stylistic and paraphrastic agreement with GPT-4o rather than an externally grounded diagnostic correctness rate. No formal equation makes the result a tautology, but the reference standard and the judge are the same system, which is a self-definitional evaluation loop. The paper’s closed-ended experiments on four unseen human-labeled datasets (Table 4) provide independent evidence of classification ability and keep the work from being wholly circular. I found no load-bearing self-citation, no imported uniqueness theorem, and no ansatz smuggled in by citation. Separately, the abstract’s “86.2% accuracy” is a mislabel: Table 3’s All* column shows BERTScore 86.2 but Accuracy 70.6; this is a reporting error, not a circular derivation, though it compounds the overstatement.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claims depend on synthetic QA generation and an LLM-based judge rather than on new physical or mathematical postulates. No invented entities are introduced. The main unvalidated inputs are the clinical correctness of the GPT-4o-generated QA pairs and the accuracy of the GPT-4o-based grading.

free parameters (5)

LoRA rank r = 8
Selected by ablation over {4, 8, 16}; r=8 gives the best BERTScore/METEOR (Table 7).
Learning rate = 2e-5
Chosen for AdamW in Sec 4.1; no sensitivity analysis reported.
Training epochs = 50
Fixed in Sec 4.1; no early stopping criterion reported.
Audio segment length = 5 seconds
Random 5-second clips sampled during training (Sec 4.1); impacts localization-type questions.
Augmentation set = 4 transformations
5dB gain, amplitude normalization, low-pass 300Hz, high-pass 3000Hz applied with equal probability (Sec 4.1).

assumptions (4)

domain assumption Source dataset labels (ICBHI, KAUH, CirCor, ZCH, SPR) are accurate ground truth for the medical conditions and sound attributes used to generate QA pairs.
Training and evaluation treat these public annotations as correct; no clinician re-adjudication is performed (Sec 3.2).
ad hoc to paper GPT-4o-generated QA pairs are clinically valid, unambiguous, and answerable from the audio alone.
QA pairs are synthesized from metadata with a prompt forbidding external assumptions (Appendix C), but the outputs are never checked by experts.
domain assumption A random 5-second audio segment contains sufficient acoustic information to answer questions about murmur grades, crackle localization, and diagnoses.
The training pipeline samples 5-second clips (Sec 4.1), yet many questions ask for localized or graded findings that may not be present in every segment.
domain assumption BERTScore and GPT-4o binary judgments are valid proxies for diagnostic correctness.
No human evaluation or clinician agreement is reported; METEOR and BERTScore are lexical/semantic similarity metrics, and the GPT-4o judge is uncalibrated (Sec 3.3.1).

how reviews work

0 comments

Cite this review

Pith. "Pith review of CaReAQA: A Cardiac and Respiratory Audio Question Answering Model for Open-Ended Diagnostic Reasoning." pith.science (2026). https://pith.science/paper/MLYPBCDR

@misc{pith2026250501199,
  author       = {Pith},
  title        = {Pith review of: CaReAQA: A Cardiac and Respiratory Audio Question Answering Model for Open-Ended Diagnostic Reasoning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MLYPBCDR}},
  note         = {Machine review of arXiv:2505.01199}
}

read the original abstract

Medical audio signals, such as heart and lung sounds, play a crucial role in clinical diagnosis. However, analyzing these signals remains challenging: traditional methods rely on handcrafted features or supervised deep learning models that demand extensive labeled datasets, limiting their scalability and applicability. To address these issues, we propose CaReAQA, an audio-language model that integrates a foundation audio model with the reasoning capabilities of large language models, enabling clinically relevant, open-ended diagnostic responses. Alongside CaReAQA, we introduce CaReSound, a benchmark dataset of annotated medical audio recordings enriched with metadata and paired question-answer examples, intended to drive progress in diagnostic reasoning research. Evaluation results show that CaReAQA achieves 86.2% accuracy on open-ended diagnostic reasoning tasks, outperforming baseline models. It also generalizes well to closed-ended classification tasks, achieving an average accuracy of 56.9% on unseen datasets. Our findings show how audio-language integration and reasoning advances medical diagnostics, enabling efficient AI systems for clinical decision support.

Figures

Figures reproduced from arXiv: 2505.01199 by the authors.

**Figure 1.** Spectrograms(left) and waveforms(right) paired with diagnostic question-and-answer outputs, demonstrating the model’s ability to analyze and answer questions about normal and abnormal heart and respiratory sounds. 1. Introduction Medical audio data, including cardiac and respiratory sounds, is a rich source of physiological information and plays a crucial role in diagnosing a wide range of health conditions. For ex… view at source ↗

**Figure 2.** Overview of CaReAQA. Our multimodal framework combining audio and text data for clinical sound analysis, enabling symptom classification, disease diagnosis, and diagnostic reasoning with a large language model. respiratory sounds, such as wheezes or crackles, are often early indicators of conditions like asthma, chronic obstructive pulmonary disease (COPD), or pulmonary fibrosis. Analyzing these sounds more effecti… view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

RA-QA: A Benchmarking System for Respiratory Audio Question Answering Under Real-World Heterogeneity
cs.SD 2026-02 conditional novelty 6.0 of 10

RA-QA converts 11 public respiratory-audio datasets into 9M template-generated QA pairs and shows current audio-language models score near zero on clinical task accuracy.

Reference graph

Works this paper leans on

26 extracted references · 12 canonical work pages · cited by 1 Pith paper

[1]

Flusense: a contactless syndromic surveillance platform for influenza-like illness in hospital waiting areas

Forsad Al Hossain, Andrew A Lover, George A Corey, Nicholas G Reich, and Tauhidur Rahman. Flusense: a contactless syndromic surveillance platform for influenza-like illness in hospital waiting areas. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies, 4 0 (1): 0 1--28, 2020

work page 2020
[2]

BUET Multi-disease Heart Sound Dataset: A Comprehensive Auscultation Dataset for Developing Computer-Aided Diagnostic Systems

Shams Nafisa Ali, Afia Zahin, Samiul Based Shuvo, Nusrat Binta Nizam, Shoyad Ibn Sabur Khan Nuhash, Sayeed Sajjad Razin, SM Sani, Farihin Rahman, Nawshad Binta Nizam, Farhat Binte Azam, et al. Buet multi-disease heart sound dataset: A comprehensive auscultation dataset for developing computer-aided diagnostic systems. arXiv preprint arXiv:2409.00724, 2024

work page Pith review arXiv 2024
[3]

Smollm2 - with great data, comes great performance, 2024

Loubna Ben Allal, Anton Lozhkov, Elie Bakouch, Gabriel Martín Blázquez, Lewis Tunstall, Agustín Piqueres, Andres Marafioti, Cyril Zakka, Leandro von Werra, and Thomas Wolf. Smollm2 - with great data, comes great performance, 2024

2024
[4]

Respiratorydatabase@tr (copd severity analysis)

Gokhan ALTAN and Yakup KUTLU. Respiratorydatabase@tr (copd severity analysis). Mendeley Data, V1, 2020

work page 2020
[5]

Meteor: An automatic metric for mt evaluation with improved correlation with human judgments

Satanjeev Banerjee and Alon Lavie. Meteor: An automatic metric for mt evaluation with improved correlation with human judgments. In Proceedings of the acl workshop on intrinsic and extrinsic evaluation measures for machine translation and/or summarization, pages 65--72, 2005

2005
[6]

Coswara: A respiratory sounds and symptoms dataset for remote screening of sars-cov-2 infection, 2023

Debarpan Bhattacharya, Neeraj Kumar Sharma, Debottam Dutta, Srikanth Raj Chetupalli, Pravin Mote, Sriram Ganapathy, Chandrakiran C, Sahiti Nori, Suhail K K, Sadhana Gonuguntla, and Murali Alagesan. Coswara: A respiratory sounds and symptoms dataset for remote screening of sars-cov-2 infection, 2023

work page 2023
[7]

Qwen-audio: Advancing universal audio understanding via unified large-scale audio-language models

Yunfei Chu, Jin Xu, Xiaohuan Zhou, Qian Yang, Shiliang Zhang, Zhijie Yan, Chang Zhou, and Jingren Zhou. Qwen-audio: Advancing universal audio understanding via unified large-scale audio-language models. arXiv preprint arXiv:2311.07919, 2023

arXiv 2023
[8]

Pengi: An audio language model for audio tasks

Soham Deshmukh, Benjamin Elizalde, Rita Singh, and Huaming Wang. Pengi: An audio language model for audio tasks. Advances in Neural Information Processing Systems, 36: 0 18090--18108, 2023

2023

Show all 26 references

[9]

Ibnian, and Mohanad Alkhodari

Luay Fraiwan, Omnia Hassanin, Mohammad Fraiwan, Basheer Khassawneh, Ali M. Ibnian, and Mohanad Alkhodari. Automatic identification of respiratory diseases from stethoscopic lung sound signals using ensemble classifiers. Biocybernetics and Biomedical Engineering, 41 0 (1): 0 1-...

2021 doi
[10]

Gemini: A family of highly capable multimodal models, 2024

Gemini. Gemini: A family of highly capable multimodal models, 2024

2024
[11]

Gemma: Open models based on gemini research and technology, 2024

Gemma Team . Gemma: Open models based on gemini research and technology, 2024

2024
[12]

Audio set: An ontology and human-labeled dataset for audio events

Jort F Gemmeke, Daniel PW Ellis, Dylan Freedman, Aren Jansen, Wade Lawrence, R Channing Moore, Manoj Plakal, and Marvin Ritter. Audio set: An ontology and human-labeled dataset for audio events. In 2017 IEEE international conference on acoustics, speech and signal processing (...

2017
[13]

Gama: A large audio-language model with advanced audio understanding and complex reasoning abilities

Sreyan Ghosh, Sonal Kumar, Ashish Seth, Chandra Kiran Reddy Evuru, Utkarsh Tyagi, S Sakshi, Oriol Nieto, Ramani Duraiswami, and Dinesh Manocha. Gama: A large audio-language model with advanced audio understanding and complex reasoning abilities. arXiv preprint arXiv:2406.11768, 2024

2024 arXiv
[14]

Listen, think, and understand

Yuan Gong, Hongyin Luo, Alexander H Liu, Leonid Karlinsky, and James Glass. Listen, think, and understand. arXiv preprint arXiv:2305.10790, 2023

2023 arXiv
[15]

Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning

Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948, 2025

2025 arXiv
[16]

Audiogpt: Understanding and generating speech, music, sound, and talking head

Rongjie Huang, Mingze Li, Dongchao Yang, Jiatong Shi, Xuankai Chang, Zhenhui Ye, Yuning Wu, Zhiqing Hong, Jiawei Huang, Jinglin Liu, et al. Audiogpt: Understanding and generating speech, music, sound, and talking head. In Proceedings of the AAAI Conference on Artificial Intell...

2024
[17]

Zchsound: open-source zju paediatric heart sound database with congenital heart disease

Weijie Jia, Yunyan Wang, Renwei Chen, Jingjing Ye, Die Li, Fei Yin, Jin Yu, Jiajia Chen, Qiang Shu, and Weize Xu. Zchsound: open-source zju paediatric heart sound database with congenital heart disease. IEEE Transactions on Biomedical Engineering, 71 0 (8): 0 2278--2286, 2024

2024
[18]

The circor digiscope dataset: from murmur detection to murmur classification

Jorge Oliveira, Francesco Renna, Paulo Dias Costa, Marcelo Nogueira, Cristina Oliveira, Carlos Ferreira, Al \' pio Jorge, Sandra Mattos, Thamine Hatem, Thiago Tavares, et al. The circor digiscope dataset: from murmur detection to murmur classification. IEEE journal of biomedic...

2021
[19]

Gpt-4 technical report, 2024

OpenAI. Gpt-4 technical report, 2024

2024
[20]

ICBHI 2017 challenge , 2023

Zhiqiang Sun. ICBHI 2017 challenge , 2023

2017
[21]

Llama: Open and efficient foundation language models

Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971, 2023

2023 arXiv
[22]

Open-ended medical visual question answering through prefix tuning of language models

Tom Van Sonsbeek, Mohammad Mahdi Derakhshani, Ivona Najdenkoska, Cees GM Snoek, and Marcel Worring. Open-ended medical visual question answering through prefix tuning of language models. In International Conference on Medical Image Computing and Computer-Assisted Intervention,...

2023
[23]

Sprsound: Open-source sjtu paediatric respiratory sound database

Qing Zhang, Jing Zhang, Jiajun Yuan, Huajie Huang, Yuhang Zhang, Baoqin Zhang, Gaomei Lv, Shuzhu Lin, Na Wang, Xin Liu, et al. Sprsound: Open-source sjtu paediatric respiratory sound database. IEEE Transactions on Biomedical Circuits and Systems, 16 0 (5): 0 867--881, 2022

2022
[24]

Bertscore: Evaluating text generation with bert

Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q Weinberger, and Yoav Artzi. Bertscore: Evaluating text generation with bert. arXiv preprint arXiv:1904.09675, 2019

1904 arXiv
[25]

Towards open respiratory acoustic foundation models: Pretraining and benchmarking

Yuwei Zhang, Tong Xia, Jing Han, Yu Yvonne Wu, Georgios Rizos, Yang Liu, Mohammed Mosuily, Jagmohan Chauhan, and Cecilia Mascolo. Towards open respiratory acoustic foundation models: Pretraining and benchmarking. In A. Globerson, L. Mackey, D. Belgrave, A. Fan, U. Paquet, J. T...

2024
[26]

Respllm: Unifying audio and text with multimodal llms for generalized respiratory health prediction

Yuwei Zhang, Tong Xia, Aaqib Saeed, and Cecilia Mascolo. Respllm: Unifying audio and text with multimodal llms for generalized respiratory health prediction. arXiv preprint arXiv:2410.05361, 2024 b

2024 arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [1]

Flusense: a contactless syndromic surveillance platform for influenza-like illness in hospital waiting areas

Forsad Al Hossain, Andrew A Lover, George A Corey, Nicholas G Reich, and Tauhidur Rahman. Flusense: a contactless syndromic surveillance platform for influenza-like illness in hospital waiting areas. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies, 4 0 (1): 0 1--28, 2020

work page 2020

[2] [2]

BUET Multi-disease Heart Sound Dataset: A Comprehensive Auscultation Dataset for Developing Computer-Aided Diagnostic Systems

Shams Nafisa Ali, Afia Zahin, Samiul Based Shuvo, Nusrat Binta Nizam, Shoyad Ibn Sabur Khan Nuhash, Sayeed Sajjad Razin, SM Sani, Farihin Rahman, Nawshad Binta Nizam, Farhat Binte Azam, et al. Buet multi-disease heart sound dataset: A comprehensive auscultation dataset for developing computer-aided diagnostic systems. arXiv preprint arXiv:2409.00724, 2024

work page Pith review arXiv 2024

[3] [3]

Smollm2 - with great data, comes great performance, 2024

Loubna Ben Allal, Anton Lozhkov, Elie Bakouch, Gabriel Martín Blázquez, Lewis Tunstall, Agustín Piqueres, Andres Marafioti, Cyril Zakka, Leandro von Werra, and Thomas Wolf. Smollm2 - with great data, comes great performance, 2024

2024

[4] [4]

Respiratorydatabase@tr (copd severity analysis)

Gokhan ALTAN and Yakup KUTLU. Respiratorydatabase@tr (copd severity analysis). Mendeley Data, V1, 2020

work page 2020

[5] [5]

Meteor: An automatic metric for mt evaluation with improved correlation with human judgments

Satanjeev Banerjee and Alon Lavie. Meteor: An automatic metric for mt evaluation with improved correlation with human judgments. In Proceedings of the acl workshop on intrinsic and extrinsic evaluation measures for machine translation and/or summarization, pages 65--72, 2005

2005

[6] [6]

Coswara: A respiratory sounds and symptoms dataset for remote screening of sars-cov-2 infection, 2023

Debarpan Bhattacharya, Neeraj Kumar Sharma, Debottam Dutta, Srikanth Raj Chetupalli, Pravin Mote, Sriram Ganapathy, Chandrakiran C, Sahiti Nori, Suhail K K, Sadhana Gonuguntla, and Murali Alagesan. Coswara: A respiratory sounds and symptoms dataset for remote screening of sars-cov-2 infection, 2023

work page 2023

[7] [7]

Qwen-audio: Advancing universal audio understanding via unified large-scale audio-language models

Yunfei Chu, Jin Xu, Xiaohuan Zhou, Qian Yang, Shiliang Zhang, Zhijie Yan, Chang Zhou, and Jingren Zhou. Qwen-audio: Advancing universal audio understanding via unified large-scale audio-language models. arXiv preprint arXiv:2311.07919, 2023

arXiv 2023

[8] [8]

Pengi: An audio language model for audio tasks

Soham Deshmukh, Benjamin Elizalde, Rita Singh, and Huaming Wang. Pengi: An audio language model for audio tasks. Advances in Neural Information Processing Systems, 36: 0 18090--18108, 2023

2023