Pith. sign in

REVIEW 4 major objections 5 minor 43 references

Comprehensive Audio Query Handling System with Integrated Expert Models and Contextual Understanding

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A modular 3.8B-parameter audio chatbot, routing queries through a BERT intent classifier and specialist models, matches 8B-parameter systems on the MMAU sound test split and beats them on timestamp questions.

desk verdict The MMAU sound result is plausible and worth a look, but the paper's custom-task headline claim is undermined by a numerical inconsistency between Tables 5 and 6. read the letter →

arxiv 2412.03980 v1 pith:EIOKFSXI submitted 2024-12-05 eess.AS

Vakada Naveen , Arvind Krishna Sridhar , Yinyi Guo , Erik Visser This is my paper · ORCID
classification eess.AS
keywords audioquerychatbotintentclassificationcontextdetectiontemporalquestionansweringtimestampreasoninglargelanguagemodelsMMAUbenchmarkretrieval-augmentedgeneration
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

This paper argues that a small, modular audio chatbot can beat much larger end-to-end audio-language models by splitting the work: a BERT router classifies the user's intent, specialized audio models answer the parts they are good at, and a 3.8B LLM assembles the final answer with the help of a machine-generated list of audio events and their timestamps. The authors report 50.75% accuracy on the MMAU sound test set, matching an 8B-parameter baseline and beating several 7B audio-language models, and 37.57% on their new timestamp-question-answering benchmark versus 30.66% for Qwen-Audio. The paper's key insight is that explicit structured metadata, in JSON, turns an audio-reasoning problem into a text-reasoning problem that a small LLM can handle. A sympathetic reader would care because the architecture points toward capable audio assistants that could run on phones rather than in the cloud.

What carries the argument

The central mechanism is the metadata bridge: a small Audio Context Detection (ACD) model, built from a CNN-based audio tagger producing frame-level event probabilities, converts raw audio into a compact JSON list of events with start times, end times, durations, and event order. A 3.8B Phi-3.5 LLM then consumes that JSON together with expert-model text outputs and chat history, turning an audio-reasoning task into a structured text-reasoning task. The JSON format carries the argument because explicit fields such as end time and event order are what make timestamp and temporal questions answerable, and the paper shows this format substantially outperforms natural-language string metadata.

What would settle it

Run the same router and LLM on the MMAU sound test split with the ACD metadata block removed or filled with empty JSON; if accuracy stays near 50.75%, the paper's claim that explicit audio-event context drives the system's performance is falsified.

Watch

Extended reading notes

Core claim

The paper claims that decomposing audio understanding into routing, specialized models, and text-based reasoning over structured metadata can outperform larger end-to-end audio-language models on several benchmarks. Specifically, a BERT-based intent classifier routes queries to experts such as Whisper, Pyannote, ACRCloud, and text-to-audio models, while a 3.8B Phi-3.5 LLM consumes the expert outputs plus Audio Context Detection (ACD) metadata in JSON form to produce the answer. On the MMAU sound test split this system scores 50.75%, matching an 8B Llama-3-based strong-captioning baseline and beating several 7B audio-language models. On the authors' new ACD-timestamp-QA benchmark it scores 37.57% versus 30.66% for Qwen-Audio and 28.56% for GAMA. The metadata bridge is the decisive component: with ground-truth audio events the LLM reaches 73.66% on temporal QA under zero-shot chain-of-thought prompting, and with the JSON format 96.35% on timestamp QA; but with predicted ACD events the temporal-QA accuracy falls to 50.34%, revealing both the power and the fragility of the approach.

Load-bearing premise

The load-bearing premise is that the ACD model's predicted event list with timestamps is a reliable and sufficiently complete transcript of the audio for whatever question is asked; the paper's own numbers show that when predicted events replace ground-truth ones, temporal-QA accuracy falls from 73.66% to 50.34%.

Editorial extensions

If this is right

  • On the MMAU sound test split, a 3.8B-parameter modular system scores 50.75%, matching an 8B Llama-3-based baseline and surpassing several 7B audio-language models.
  • On the custom timestamp-QA benchmark, the system reaches 37.57% accuracy, ahead of Qwen-Audio at 30.66% and GAMA at 28.56%.
  • On the temporal-QA benchmark, the same system trails GAMA at 50.34% versus 57.53%, showing that chronological order reasoning is not yet solved by this approach.
  • BERT-based intent routing at 85% overall accuracy is a cheaper and more accurate router than few-shot LLM routing at 37%, so the routing stage does not need a large model.
  • Using JSON metadata with explicit end times, durations, and event order yields 96.35% accuracy on timestamp QA with ground-truth metadata, versus 89.75% for a natural-language string format.

Reading between the lines

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

  • Editorial inference: the same architecture could reach much higher accuracy by swapping in a stronger audio tagger, since the language model already scores 73.66% when given ground-truth events; the bottleneck is the metadata, not the LLM.
  • Editorial inference: the intent dataset and the two QA benchmarks are reusable test assets, so other audio assistants can be scored on the same temporal and timestamp questions, making the claimed advantage directly checkable.
  • Editorial inference: a privacy-conscious deployment could run the small expert models on-device and send only text metadata to the LLM, a split the paper mentions for resource reasons rather than for privacy.
  • Editorial inference: the reported advantage may be partly a property of how the benchmark questions are phrased; testing the same system on naturally occurring user audio questions, not generated QA pairs, would show whether the metadata bridge generalizes beyond the curated test sets.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes a modular audio chatbot system: a BERT-based intent classifier routes user queries to specialized expert models (ASR, speaker diarization, music identification, text-to-audio, AQA-LLM), and a 3.8B Phi-3.5 LLM forms the final response by consuming outputs from these experts together with JSON metadata from an Audio Context Detection (ACD) model. The authors introduce a new audio-intent dataset, two custom QA benchmarks (ACD-timestamp-QA and ACD-temporal-QA), and report evaluations on those benchmarks plus the MMAU sound test split. The central claims are that the BERT router beats an LLM few-shot router, that JSON metadata outperforms string metadata, and that the proposed Phi+ACD system outperforms 7B-parameter large audio-language models on some custom tasks and matches or surpasses 7B-parameter models on MMAU sound.

Significance. If the reported results held, this would be a practically valuable demonstration that a small text-only LLM, combined with off-the-shelf expert audio models and structured ACD metadata, can be competitive with much larger end-to-end audio-language models while remaining edge-deployable. The paper has strengths: it evaluates on the external MMAU benchmark, compares against several strong baselines, introduces a new intent-classification dataset, and the modular design is clearly described. However, the current evaluation contains an internal numerical inconsistency in the headline comparison, the prompting configuration is selected on the same test set used for the SOTA comparison, and the abstract overstates the MMAU result. These issues mean the central quantitative claims are not currently established.

major comments (4)
  1. [4.3.1 and Table 6] The configuration named Phi+ACD in Table 6 is not the configuration that produces the reported number. In Section 4.3.1, Table 5 reports 50.34% for Zeroshot, 48.54% for Zeroshot+CoT, and 46.12% for Fewshot+CoT when the input is predicted ACD metadata. The text states that 'assuming the ACD model predictions are reliable, Zeroshot + CoT is the best prompting method' and names this configuration Phi+ACD. Table 6 then lists Phi+ACD at 50.34% on ACD-temporal-QA, which is exactly the Zeroshot (no-CoT) value, not the 48.54% of the named configuration. If the intended configuration includes CoT, the reported value is wrong; if it is Zeroshot without CoT, the method name and selection narrative are wrong. Either way, the claimed advantage over Qwen on ACD-temporal-QA is not attached to a reproducible method.
  2. [4.3.1-4.3.2] The prompting configuration was selected on the same test set used for the headline comparison. Table 5 is computed on ACD-temporal-QA, and the best configuration is then compared against Qwen and GAMA on that same dataset in Table 6. No validation split or held-out evaluation is described, so the custom-task results are post-hoc selections and can overstate the performance of the chosen method. All accuracies are point estimates without confidence intervals or significance tests; differences such as 37.57% vs. 30.66% on ACD-timestamp-QA are not demonstrated to be significant. The custom datasets are not released, so an independent check of these numbers is impossible.
  3. [Abstract and Table 7] The abstract claims the system 'outperforms models in the 7B parameter size range on the sound testset of the MMAU benchmark,' but Table 7 reports Llama-3-Instruct + strong caption (8B) at exactly the same 50.75% as Phi-3.5 + ACD. At best this is a tie, not an outperformance. Additionally, the word 'significantly' in the abstract's claim about custom-task accuracy is unsupported because no significance testing is reported.
  4. [2.4, 4.1, Table 5] The system's practical accuracy depends on the ACD model's predicted metadata, but this dependency is only tested in-distribution. The ACD model is built on PANN CNN10 pretrained on AudioSet, and the custom QA datasets are generated from AudioSet ground-truth events. Table 5 shows that substituting predicted metadata for ground-truth metadata drops accuracy on ACD-temporal-QA from 73.66% to at most 50.34% for any predicted-metadata configuration. Because the evaluation is on the same data distribution the ACD was trained on, the reliability of the system on out-of-distribution audio, where the 'on-device deployment' claim would matter, is not established.
minor comments (5)
  1. [3.2] The text uses 'V oiceFilter' with a broken spacing; it should read 'VoiceFilter'.
  2. [Table 1] The intent class name 'Audio/Text to Audio' is ambiguous; clarifying the intended query types for this class would help reproducibility.
  3. [Table 3] The caption reports 'Overall accuracy 0.85' and '0.37', but the table body has no overall row; an explicit accuracy row or a definition of the metric should be added.
  4. [4.1] The paper alternates between 'ACD-timestamp-QA' and 'ACD-timestamps-QA'; a single consistent name should be used throughout.
  5. [Appendix D and Figure 5] The prompts are provided only as an image; machine-readable prompt templates in the appendix would support reproducibility, especially since the number of few-shot examples is not specified for the intent-classifier few-shot setting.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the headline results are anchored to an external benchmark and to empirically scored custom QA labels, not to the paper's own definitions.

full rationale

The paper's derivation chain has no step in which an output is defined as its input or a fitted parameter is renamed as a prediction. The custom ACD-timestamp-QA and ACD-temporal-QA datasets are generated from AudioSet ground-truth events via GPT-4 (Section 4.1), and the LLM answers are scored against those ground-truth labels; the system's ACD module is a separate pretrained audio tagger, not a component fitted to the test answers. The paper itself shows that the ACD metadata is not equivalent to the ground truth: with ground-truth metadata, Zeroshot+CoT gives 73.66%, whereas with predicted ACD metadata the same method gives 48.54% (Table 5), so the reported pipeline output is not equal to its input by construction. The MMAU sound-set result (50.75%, Table 7) is evaluated against an external benchmark with fixed published baselines, so it cannot reduce to the authors' own definitions. The main self-citations (AQA-LLM, Mahfuz et al.) describe components and prior context, but the headline custom-task comparisons use Phi+ACD and external LALM baselines, and no uniqueness or forcing argument is imported from those citations. The manifest limitation 'assuming the ACD model predictions are reliable' (Section 4.3.1) and the numeric inconsistency between Table 5's Zeroshot+CoT (48.54%) and Table 6's Phi+ACD (50.34%, the plain Zeroshot value) are correctness and reproducibility concerns, not circularity: the reported number is potentially mislabeled, not tautologically entailed. Hence no circular step is established by quotation.

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

The central results rest on the reliability of the ACD metadata representation and the synthetic QA benchmarks. The paper introduces no new physical or mathematical entities, but it relies on several domain assumptions about metadata sufficiency, LLM temporal reasoning, and intent routing accuracy. The only numerical free parameters explicitly disclosed are the chat-history limit and few-shot count; the ACD threshold and GPT-4 generation prompt are unstated. The post-hoc selection of the Zeroshot + CoT configuration on the test set is the most consequential free choice.

free parameters (5)
  • ACD event detection probability threshold
    The ACD model outputs frame-level probabilities; events and timestamps are determined by applying thresholds (Section 3.3). The threshold value is not reported, and the reported accuracy depends on it.
  • Chat history length limit = 10 turns
    Section 3.3 states chat history is limited to the last 10 turns because of context length and observed degradation. This is a hand-chosen constant that affects responses.
  • Number of few-shot examples = 2
    Appendix D / Table 5 says the fewshot method has 2 example QA pairs in the prompt. The number is chosen by hand.
  • GPT-4 data generation prompt
    The custom QA datasets are generated using GPT-4 data augmentation (Section 4.1); the prompt is not fully specified, so the generation process is a free choice that shapes the benchmarks.
  • Zeroshot + CoT method selection = selected after test evaluation
    The paper chooses Zeroshot + CoT as the system method because it had the highest accuracy on the ACD-temporal-QA test set (Table 5), which is a post-hoc selection on the test data.
assumptions (5)
  • domain assumption Audio content relevant to user queries is faithfully captured by a list of acoustic events with start times and durations emitted by the ACD model.
    The whole RAG design (Section 3.3) feeds only ACD metadata as audio context to the LLM; if event detection omits or mislabels the relevant audio content, the final answer is wrong.
  • domain assumption A text-only LLM (Phi-3.5) can reason correctly about timestamps, duration, and temporal order when given structured JSON metadata.
    The method assumes the LLM's arithmetic and temporal reasoning from text is sufficient; performance relies on this and is shown to be sensitive to metadata format (Table 4).
  • domain assumption The BERT intent classifier routes queries to the correct expert model with high accuracy.
    Section 3.1 reports an overall accuracy of 0.85; routing errors send the query to the wrong expert and are not measured in the end-to-end evaluations.
  • ad hoc to paper GPT-4-generated QA pairs from AudioSet ground-truth events are valid tests of real user temporal audio reasoning.
    The custom ACD-timestamp-QA and ACD-temporal-QA datasets (Section 4.1) are synthetic; their validity as benchmarks is assumed, and the distribution is the AudioSet training domain of the ACD model.
  • domain assumption ACD predictions on AudioSet-like audio are accurate enough to support the temporal QA comparisons.
    Section 4.3.1 states 'assuming the ACD model predictions are reliable'; the drop from 73.66% to 50.34% when using predictions shows this assumption is fragile.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Comprehensive Audio Query Handling System with Integrated Expert Models and Contextual Understanding." pith.science (2026). https://pith.science/paper/EIOKFSXI

@misc{pith2026241203980,
  author       = {Pith},
  title        = {Pith review of: Comprehensive Audio Query Handling System with Integrated Expert Models and Contextual Understanding},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EIOKFSXI}},
  note         = {Machine review of arXiv:2412.03980}
}
read the original abstract

This paper presents a comprehensive chatbot system designed to handle a wide range of audio-related queries by integrating multiple specialized audio processing models. The proposed system uses an intent classifier, trained on a diverse audio query dataset, to route queries about audio content to expert models such as Automatic Speech Recognition (ASR), Speaker Diarization, Music Identification, and Text-to-Audio generation. A 3.8 B LLM model then takes inputs from an Audio Context Detection (ACD) module extracting audio event information from the audio and post processes text domain outputs from the expert models to compute the final response to the user. We evaluated the system on custom audio tasks and MMAU sound set benchmarks. The custom datasets were motivated by target use cases not covered in industry benchmarks and included ACD-timestamp-QA (Question Answering) as well as ACD-temporal-QA datasets to evaluate timestamp and temporal reasoning questions, respectively. First we determined that a BERT based Intent Classifier outperforms LLM-fewshot intent classifier in routing queries. Experiments further show that our approach significantly improves accuracy on some custom tasks compared to state-of-the-art Large Audio Language Models and outperforms models in the 7B parameter size range on the sound testset of the MMAU benchmark, thereby offering an attractive option for on device deployment.

Figures

Figures reproduced from arXiv: 2412.03980 by the authors.

Figure 1
Figure 1. Proposed chatbot system Class Training Count Test Count Total Count Audio/Text to Audio 1909 478 2387 LLM 1893 473 2366 Music recommendation 917 230 1147 ASR whisper 778 194 972 Music identification 643 161 804 Speaker ID, Diarization, counting 602 150 752 Source separation/removal 243 61 304 Unsupported 2343 586 2929 Total 10328 2333 12661 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Sample queries from intent classification dataset [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Sample queries from ACD-timestamp-QA dataset [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Sample queries from ACD-temporal-QA dataset [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]
Figure 5
Figure 5. Figure 5: Prompts used for various methods [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Qualitative examples generated by the proposed chatbot [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 19 canonical work pages

  1. [1]

    Marah Abdin, Jyoti Aneja, Hany Awadalla, Ahmed Awadallah, Ammar Ahmad Awan, Nguyen Bach, Amit Bahree, Arash Bakhtiari, Jianmin Bao, Harkirat Behl, et al. 2024. Phi-3 technical report: A highly capable language model locally on your phone. arXiv preprint arXiv:2404.14219

    arXiv 2024

  2. [2]

    ACRCloud. 2024. Acrcloud: Audio recognition platform. https://www.acrcloud.com/. Available at: https://www.acrcloud.com/

    work page 2024

  3. [3]

    Inc. AWS. 2017. https://docs.aws.amazon.com/lex/latest/dg/what-is.html Amazon lex – build conversation bots . [cited 30/11/2024]

    work page 2017

  4. [4]

    Emanuele Bastianelli, Andrea Vanzo, Pawel Swietojanski, and Verena Rieser. 2020. Slurp: A spoken language understanding resource package. arXiv preprint arXiv:2011.13205

    arXiv 2020

  5. [5]

    Alberto Benayas, Sicilia Miguel- \'A ngel, and Mar c al Mora-Cantallops. 2024. Enhancing intent classifier training with large language model-generated data. Applied Artificial Intelligence, 38(1):2414483

    work page 2024

  6. [6]

    Herv \'e Bredin. 2023. pyannote. audio 2.1 speaker diarization pipeline: principle, benchmark, and recipe. In 24th INTERSPEECH Conference (INTERSPEECH 2023), pages 1983--1987. ISCA

    work page 2023

  7. [7]

    I \ n igo Casanueva, Tadas Tem c inas, Daniela Gerz, Matthew Henderson, and Ivan Vuli \'c . 2020. Efficient intent detection with dual sentence encoders. arXiv preprint arXiv:2003.04807

    arXiv 2020

  8. [8]

    Chakib Chaouch, Murtadha Arif Bin Sahbudin, Marco Scarpa, Salvatore Serrano, et al. 2020. Audio fingerprint database structure using k-modes clustering. Journal of Advanced Research in Dynamical and Control Systems, 12(4):1545--1554

    work page 2020

Show all 43 references
  1. [9]

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

    2023 arXiv

  2. [10]

    Alice Coucke, Alaa Saade, Adrien Ball, Th \'e odore Bluche, Alexandre Caulier, David Leroy, Cl \'e ment Doumouro, Thibault Gisselbrecht, Francesco Caltagirone, Thibaut Lavril, et al. 2018. Snips voice platform: an embedded spoken language understanding system for private-by-de...

    2018 arXiv

  3. [11]

    Xiang Deng, Yu Gu, Boyuan Zheng, Shijie Chen, Sam Stevens, Boshi Wang, Huan Sun, and Yu Su. 2024. Mind2web: Towards a generalist agent for the web. Advances in Neural Information Processing Systems, 36

    2024

  4. [12]

    Zhihao Du, Jiaming Wang, Qian Chen, Yunfei Chu, Zhifu Gao, Zerui Li, Kai Hu, Xiaohuan Zhou, Jin Xu, Ziyang Ma, et al. 2023. Lauragpt: Listen, attend, understand, and regenerate audio with gpt. arXiv preprint arXiv:2310.04673

    2023 arXiv

  5. [13]

    Jack FitzGerald, Christopher Hench, Charith Peris, Scott Mackie, Kay Rottmann, Ana Sanchez, Aaron Nash, Liam Urbach, Vishesh Kakarala, Richa Singh, et al. 2022. Massive: A 1m-example multilingual natural language understanding dataset with 51 typologically-diverse languages. a...

    2022 arXiv

  6. [14]

    Yingqiang Ge, Wenyue Hua, Kai Mei, Juntao Tan, Shuyuan Xu, Zelong Li, Yongfeng Zhang, et al. 2024. Openagi: When llm meets domain experts. Advances in Neural Information Processing Systems, 36

    2024

  7. [15]

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

    2017

  8. [16]

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

    2024 arXiv

  9. [17]

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

    2023 arXiv

  10. [18]

    Google. 2010. https://dialogflow.com/ Dialogflow . [cited 30/11/2024]

    2010

  11. [19]

    Charles T Hemphill, John J Godfrey, and George R Doddington. 1990. The atis spoken language systems pilot corpus. In Speech and Natural Language: Proceedings of a Workshop Held at Hidden Valley, Pennsylvania, June 24-27, 1990

    1990

  12. [20]

    Rongjie Huang, Jiawei Huang, Dongchao Yang, Yi Ren, Luping Liu, Mingze Li, Zhenhui Ye, Jinglin Liu, Xiang Yin, and Zhou Zhao. 2023. Make-an-audio: Text-to-audio generation with prompt-enhanced diffusion models. In International Conference on Machine Learning, pages 13916--13932. PMLR

    2023

  13. [21]

    Rongjie Huang, Mingze Li, Dongchao Yang, Jiatong Shi, Xuankai Chang, Zhenhui Ye, Yuning Wu, Zhiqing Hong, Jiawei Huang, Jinglin Liu, et al. 2024. Audiogpt: Understanding and generating speech, music, sound, and talking head. In Proceedings of the AAAI Conference on Artificial ...

    2024

  14. [22]

    Geunwoo Kim, Pierre Baldi, and Stephen McAleer. 2024. Language models can solve computer tasks. Advances in Neural Information Processing Systems, 36

    2024

  15. [23]

    Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa. 2022. Large language models are zero-shot reasoners. Advances in neural information processing systems, 35:22199--22213

    2022

  16. [24]

    Qiuqiang Kong, Yin Cao, Turab Iqbal, Yuxuan Wang, Wenwu Wang, and Mark D Plumbley. 2020. Panns: Large-scale pretrained audio neural networks for audio pattern recognition. IEEE/ACM Transactions on Audio, Speech, and Language Processing, 28:2880--2894

    2020

  17. [25]

    Vimal Kumar, Priyam Srivastava, Ashay Dwivedi, Ishan Budhiraja, Debjani Ghosh, Vikas Goyal, and Ruchika Arora. 2023. Large-language-models (llm)-based ai chatbots: Architecture, in-depth analysis and their performance evaluation. In International Conference on Recent Trends in...

    2023

  18. [26]

    Stefan Larson, Anish Mahendran, Joseph J Peper, Christopher Clarke, Andrew Lee, Parker Hill, Jonathan K Kummerfeld, Kevin Leach, Michael A Laurenzano, Lingjia Tang, et al. 2019. An evaluation dataset for intent classification and out-of-scope prediction. arXiv preprint arXiv:1...

    2019 arXiv

  19. [27]

    Jinhua Liang, Huan Zhang, Haohe Liu, Yin Cao, Qiuqiang Kong, Xubo Liu, Wenwu Wang, Mark D Plumbley, Huy Phan, and Emmanouil Benetos. 2024. Wavcraft: Audio editing and generation with natural language prompts. arXiv preprint arXiv:2403.09527

    2024 arXiv

  20. [28]

    Lefteris Loukas, Ilias Stogiannidis, Odysseas Diamantopoulos, Prodromos Malakasiotis, and Stavros Vassos. 2023. Making llms worth every penny: Resource-limited text classification in banking. In Proceedings of the Fourth ACM International Conference on AI in Finance, pages 392--400

    2023

  21. [29]

    Rehana Mahfuz, Yinyi Guo, and Erik Visser. 2023. Improving audio captioning using semantic similarity metrics. In ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 1--5. IEEE

    2023

  22. [30]

    Mishaim Malik, Muhammad Kamran Malik, Khawar Mehmood, and Imran Makhdoom. 2021. Automatic speech recognition: a survey. Multimedia Tools and Applications, 80:9411--9457

    2021

  23. [31]

    Microsoft. 2015. https://www.luis.ai/home Microsoft cognitive services: Luis . [cited 30/11/2024]

    2015

  24. [32]

    Soham Parikh, Quaizar Vohra, Prashil Tumbade, and Mitul Tiwari. 2023. Exploring zero and few-shot techniques for intent classification. arXiv preprint arXiv:2305.07157

    2023 arXiv

  25. [33]

    Tae Jin Park, Naoyuki Kanda, Dimitrios Dimitriadis, Kyu J Han, Shinji Watanabe, and Shrikanth Narayanan. 2022. A review of speaker diarization: Recent advances with deep learning. Computer Speech & Language, 72:101317

    2022

  26. [34]

    Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, and Ilya Sutskever. 2023. Robust speech recognition via large-scale weak supervision. In International conference on machine learning, pages 28492--28518. PMLR

    2023

  27. [35]

    S Sakshi, Utkarsh Tyagi, Sonal Kumar, Ashish Seth, Ramaneswaran Selvakumar, Oriol Nieto, Ramani Duraiswami, Sreyan Ghosh, and Dinesh Manocha. 2024. Mmau: A massive multi-task audio understanding and reasoning benchmark. arXiv preprint arXiv:2410.19168

    2024 arXiv

  28. [36]

    Arvind Krishna Sridhar, Yinyi Guo, and Erik Visser. 2024. Enhancing temporal understanding in audio question answering for large audio language models. arXiv preprint arXiv:2409.06223

    2024 arXiv

  29. [37]

    Dídac Surís, Sachit Menon, and Carl Vondrick. 2023. Vipergpt: Visual inference via python execution for reasoning. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 11888--11898

    2023

  30. [38]

    Changli Tang, Wenyi Yu, Guangzhi Sun, Xianzhao Chen, Tian Tan, Wei Li, Lu Lu, Zejun Ma, and Chao Zhang. 2023. Salmonn: Towards generic hearing abilities for large language models. arXiv preprint arXiv:2310.13289

    2023 arXiv

  31. [39]

    Quan Wang, Hannah Muckenhirn, Kevin Wilson, Prashant Sridhar, Zelin Wu, John Hershey, Rif A Saurous, Ron J Weiss, Ye Jia, and Ignacio Lopez Moreno. 2018. Voicefilter: Targeted voice separation by speaker-conditioned spectrogram masking. arXiv preprint arXiv:1810.04826

    2018 arXiv

  32. [40]

    Jian Wu, Yashesh Gaur, Zhuo Chen, Long Zhou, Yimeng Zhu, Tianrui Wang, Jinyu Li, Shujie Liu, Bo Ren, Linquan Liu, et al. 2023. On decoder-only architecture for speech-to-text and large language model integration. In 2023 IEEE Automatic Speech Recognition and Understanding Work...

    2023

  33. [41]

    Yuanjun Zhao, Xianjun Xia, and Roberto Togneri. 2019. Applications of deep learning to audio generation. IEEE Circuits and Systems Magazine, 19(4):19--38

    2019

  34. [42]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list...

  35. [43]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.