MEERKAT: Audio-Visual Large Language Model for Grounding in Space and Time

Sanjoy Chowdhury; Sayan Nag; Subhrajyoti Dasgupta; Jun Chen; Mohamed Elhoseiny; Ruohan Gao; Dinesh Manocha

doi:10.1007/978-3-031-73039-9_4

MEERKAT: Audio-Visual Large Language Model for Grounding in Space and Time

Sanjoy Chowdhury^*, Sayan Nag^*, Subhrajyoti Dasgupta^*, Jun Chen, Mohamed Elhoseiny, Ruohan Gao, Dinesh Manocha

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Leveraging Large Language Models’ remarkable proficiency in text-based tasks, recent works on Multi-modal LLMs (MLLMs) extend them to other modalities like vision and audio. However, the progress in these directions has been mostly focused on tasks that only require a coarse-grained understanding of the audio-visual semantics. We present Meerkat, an audio-visual LLM equipped with a fine-grained understanding of image and audio both spatially and temporally. With a new modality alignment module based on optimal transport and a cross-attention module that enforces audio-visual consistency, Meerkat can tackle challenging tasks such as audio referred image grounding, image guided audio temporal localization, and audio-visual fact-checking. Moreover, we carefully curate a large dataset AVFIT that comprises 3M instruction tuning samples collected from open-source datasets, and introduce MeerkatBench that unifies five challenging audio-visual tasks. We achieve state-of-the-art performance on all these downstream tasks with a relative improvement of up to 37.12%.

Original language	English (US)
Title of host publication	Computer Vision – ECCV 2024 - 18th European Conference, Proceedings
Editors	Aleš Leonardis, Elisa Ricci, Stefan Roth, Olga Russakovsky, Torsten Sattler, Gül Varol
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	52-70
Number of pages	19
ISBN (Print)	9783031730382
DOIs	https://doi.org/10.1007/978-3-031-73039-9_4
State	Published - 2025
Event	18th European Conference on Computer Vision, ECCV 2024 - Milan, Italy Duration: Sep 29 2024 → Oct 4 2024

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	15122 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	18th European Conference on Computer Vision, ECCV 2024
Country/Territory	Italy
City	Milan
Period	09/29/24 → 10/4/24

Bibliographical note

Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

Keywords

Audio-Visual LLM
AV Localization
AVFIT Dataset

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-031-73039-9_4

Cite this

Chowdhury, S., Nag, S., Dasgupta, S., Chen, J., Elhoseiny, M., Gao, R., & Manocha, D. (2025). MEERKAT: Audio-Visual Large Language Model for Grounding in Space and Time. In A. Leonardis, E. Ricci, S. Roth, O. Russakovsky, T. Sattler, & G. Varol (Eds.), Computer Vision – ECCV 2024 - 18th European Conference, Proceedings (pp. 52-70). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 15122 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-73039-9_4

Chowdhury, Sanjoy ; Nag, Sayan ; Dasgupta, Subhrajyoti et al. / MEERKAT : Audio-Visual Large Language Model for Grounding in Space and Time. Computer Vision – ECCV 2024 - 18th European Conference, Proceedings. editor / Aleš Leonardis ; Elisa Ricci ; Stefan Roth ; Olga Russakovsky ; Torsten Sattler ; Gül Varol. Springer Science and Business Media Deutschland GmbH, 2025. pp. 52-70 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{c8683d530599430baa57440f3e81bcc4,

title = "MEERKAT: Audio-Visual Large Language Model for Grounding in Space and Time",

abstract = "Leveraging Large Language Models{\textquoteright} remarkable proficiency in text-based tasks, recent works on Multi-modal LLMs (MLLMs) extend them to other modalities like vision and audio. However, the progress in these directions has been mostly focused on tasks that only require a coarse-grained understanding of the audio-visual semantics. We present Meerkat, an audio-visual LLM equipped with a fine-grained understanding of image and audio both spatially and temporally. With a new modality alignment module based on optimal transport and a cross-attention module that enforces audio-visual consistency, Meerkat can tackle challenging tasks such as audio referred image grounding, image guided audio temporal localization, and audio-visual fact-checking. Moreover, we carefully curate a large dataset AVFIT that comprises 3M instruction tuning samples collected from open-source datasets, and introduce MeerkatBench that unifies five challenging audio-visual tasks. We achieve state-of-the-art performance on all these downstream tasks with a relative improvement of up to 37.12%.",

keywords = "Audio-Visual LLM, AV Localization, AVFIT Dataset",

author = "Sanjoy Chowdhury and Sayan Nag and Subhrajyoti Dasgupta and Jun Chen and Mohamed Elhoseiny and Ruohan Gao and Dinesh Manocha",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.; 18th European Conference on Computer Vision, ECCV 2024 ; Conference date: 29-09-2024 Through 04-10-2024",

year = "2025",

doi = "10.1007/978-3-031-73039-9_4",

language = "English (US)",

isbn = "9783031730382",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "52--70",

editor = "Ale{\v s} Leonardis and Elisa Ricci and Stefan Roth and Olga Russakovsky and Torsten Sattler and G{\"u}l Varol",

booktitle = "Computer Vision – ECCV 2024 - 18th European Conference, Proceedings",

address = "Germany",

}

Chowdhury, S, Nag, S, Dasgupta, S, Chen, J , Elhoseiny, M, Gao, R & Manocha, D 2025, MEERKAT: Audio-Visual Large Language Model for Grounding in Space and Time. in A Leonardis, E Ricci, S Roth, O Russakovsky, T Sattler & G Varol (eds), Computer Vision – ECCV 2024 - 18th European Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 15122 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 52-70, 18th European Conference on Computer Vision, ECCV 2024, Milan, Italy, 09/29/24. https://doi.org/10.1007/978-3-031-73039-9_4

MEERKAT: Audio-Visual Large Language Model for Grounding in Space and Time. / Chowdhury, Sanjoy; Nag, Sayan; Dasgupta, Subhrajyoti et al.
Computer Vision – ECCV 2024 - 18th European Conference, Proceedings. ed. / Aleš Leonardis; Elisa Ricci; Stefan Roth; Olga Russakovsky; Torsten Sattler; Gül Varol. Springer Science and Business Media Deutschland GmbH, 2025. p. 52-70 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 15122 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - MEERKAT

T2 - 18th European Conference on Computer Vision, ECCV 2024

AU - Chowdhury, Sanjoy

AU - Nag, Sayan

AU - Dasgupta, Subhrajyoti

AU - Chen, Jun

AU - Elhoseiny, Mohamed

AU - Gao, Ruohan

AU - Manocha, Dinesh

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

PY - 2025

Y1 - 2025

N2 - Leveraging Large Language Models’ remarkable proficiency in text-based tasks, recent works on Multi-modal LLMs (MLLMs) extend them to other modalities like vision and audio. However, the progress in these directions has been mostly focused on tasks that only require a coarse-grained understanding of the audio-visual semantics. We present Meerkat, an audio-visual LLM equipped with a fine-grained understanding of image and audio both spatially and temporally. With a new modality alignment module based on optimal transport and a cross-attention module that enforces audio-visual consistency, Meerkat can tackle challenging tasks such as audio referred image grounding, image guided audio temporal localization, and audio-visual fact-checking. Moreover, we carefully curate a large dataset AVFIT that comprises 3M instruction tuning samples collected from open-source datasets, and introduce MeerkatBench that unifies five challenging audio-visual tasks. We achieve state-of-the-art performance on all these downstream tasks with a relative improvement of up to 37.12%.

AB - Leveraging Large Language Models’ remarkable proficiency in text-based tasks, recent works on Multi-modal LLMs (MLLMs) extend them to other modalities like vision and audio. However, the progress in these directions has been mostly focused on tasks that only require a coarse-grained understanding of the audio-visual semantics. We present Meerkat, an audio-visual LLM equipped with a fine-grained understanding of image and audio both spatially and temporally. With a new modality alignment module based on optimal transport and a cross-attention module that enforces audio-visual consistency, Meerkat can tackle challenging tasks such as audio referred image grounding, image guided audio temporal localization, and audio-visual fact-checking. Moreover, we carefully curate a large dataset AVFIT that comprises 3M instruction tuning samples collected from open-source datasets, and introduce MeerkatBench that unifies five challenging audio-visual tasks. We achieve state-of-the-art performance on all these downstream tasks with a relative improvement of up to 37.12%.

KW - Audio-Visual LLM

KW - AV Localization

KW - AVFIT Dataset

UR - http://www.scopus.com/inward/record.url?scp=85208596916&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-73039-9_4

DO - 10.1007/978-3-031-73039-9_4

M3 - Conference contribution

AN - SCOPUS:85208596916

SN - 9783031730382

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 52

EP - 70

BT - Computer Vision – ECCV 2024 - 18th European Conference, Proceedings

A2 - Leonardis, Aleš

A2 - Ricci, Elisa

A2 - Roth, Stefan

A2 - Russakovsky, Olga

A2 - Sattler, Torsten

A2 - Varol, Gül

PB - Springer Science and Business Media Deutschland GmbH

Y2 - 29 September 2024 through 4 October 2024

ER -

Chowdhury S, Nag S, Dasgupta S, Chen J , Elhoseiny M, Gao R et al. MEERKAT: Audio-Visual Large Language Model for Grounding in Space and Time. In Leonardis A, Ricci E, Roth S, Russakovsky O, Sattler T, Varol G, editors, Computer Vision – ECCV 2024 - 18th European Conference, Proceedings. Springer Science and Business Media Deutschland GmbH. 2025. p. 52-70. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-73039-9_4