@inproceedings{huang-etal-2026-masked,
title = "Masked Text-to-Audio Flow-Matching and Reward Feedback Optimization",
author = "Huang, Rongjie and
Yang, Dongchao and
Guo, Wenxiang and
Liu, Huadai and
Cheng, Xize and
Wang, Zehan and
Zhao, Zhou and
Wu, Xixin and
Meng, Helen M.",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.1891/",
pages = "37953--37966",
ISBN = "979-8-89176-395-1",
abstract = "Flow-matching generative models have created significant milestones in text-to-audio generation, powered by scalable training with increased data, computational resources, and model size, while their scalable inference remains less explored. In this work, we propose MaskAudioFlow, a continuous flow-matching transformer with masked generative modeling designed for scaling text-to-audio inference-time prediction. Specifically, MaskAudioFlow 1) masks spans of audio frames in training and approximates the continuous velocity vector field with flow-matching objective, and 2) performs inference via masked prediction, where we mask out generation and re-predict them through iterative decoding. To reduce the gap between generation and human preferences, we fine-tune MaskAudioFlow using reward signals from text-audio correspondence and perceptual aesthetics. Experimental results demonstrate that MaskAudioFlow achieves state-of-the-art performance in text-to-audio generation, effectively scaling inference-time computation through iterative masked prediction. Moreover, the preference-tuned model demonstrates superior text-audio alignment faithfulness and enhanced perceptual aesthetics. Audio samples are available at https://MaskAudio.github.io"
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="huang-etal-2026-masked">
<titleInfo>
<title>Masked Text-to-Audio Flow-Matching and Reward Feedback Optimization</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rongjie</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dongchao</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wenxiang</namePart>
<namePart type="family">Guo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Huadai</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xize</namePart>
<namePart type="family">Cheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zehan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhou</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xixin</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Helen</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Meng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>Flow-matching generative models have created significant milestones in text-to-audio generation, powered by scalable training with increased data, computational resources, and model size, while their scalable inference remains less explored. In this work, we propose MaskAudioFlow, a continuous flow-matching transformer with masked generative modeling designed for scaling text-to-audio inference-time prediction. Specifically, MaskAudioFlow 1) masks spans of audio frames in training and approximates the continuous velocity vector field with flow-matching objective, and 2) performs inference via masked prediction, where we mask out generation and re-predict them through iterative decoding. To reduce the gap between generation and human preferences, we fine-tune MaskAudioFlow using reward signals from text-audio correspondence and perceptual aesthetics. Experimental results demonstrate that MaskAudioFlow achieves state-of-the-art performance in text-to-audio generation, effectively scaling inference-time computation through iterative masked prediction. Moreover, the preference-tuned model demonstrates superior text-audio alignment faithfulness and enhanced perceptual aesthetics. Audio samples are available at https://MaskAudio.github.io</abstract>
<identifier type="citekey">huang-etal-2026-masked</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.1891/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>37953</start>
<end>37966</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Masked Text-to-Audio Flow-Matching and Reward Feedback Optimization
%A Huang, Rongjie
%A Yang, Dongchao
%A Guo, Wenxiang
%A Liu, Huadai
%A Cheng, Xize
%A Wang, Zehan
%A Zhao, Zhou
%A Wu, Xixin
%A Meng, Helen M.
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F huang-etal-2026-masked
%X Flow-matching generative models have created significant milestones in text-to-audio generation, powered by scalable training with increased data, computational resources, and model size, while their scalable inference remains less explored. In this work, we propose MaskAudioFlow, a continuous flow-matching transformer with masked generative modeling designed for scaling text-to-audio inference-time prediction. Specifically, MaskAudioFlow 1) masks spans of audio frames in training and approximates the continuous velocity vector field with flow-matching objective, and 2) performs inference via masked prediction, where we mask out generation and re-predict them through iterative decoding. To reduce the gap between generation and human preferences, we fine-tune MaskAudioFlow using reward signals from text-audio correspondence and perceptual aesthetics. Experimental results demonstrate that MaskAudioFlow achieves state-of-the-art performance in text-to-audio generation, effectively scaling inference-time computation through iterative masked prediction. Moreover, the preference-tuned model demonstrates superior text-audio alignment faithfulness and enhanced perceptual aesthetics. Audio samples are available at https://MaskAudio.github.io
%U https://aclanthology.org/2026.findings-acl.1891/
%P 37953-37966
Markdown (Informal)
[Masked Text-to-Audio Flow-Matching and Reward Feedback Optimization](https://aclanthology.org/2026.findings-acl.1891/) (Huang et al., Findings 2026)
ACL
- Rongjie Huang, Dongchao Yang, Wenxiang Guo, Huadai Liu, Xize Cheng, Zehan Wang, Zhou Zhao, Xixin Wu, and Helen M. Meng. 2026. Masked Text-to-Audio Flow-Matching and Reward Feedback Optimization. In Findings of the Association for Computational Linguistics: ACL 2026, pages 37953–37966, San Diego, California, United States. Association for Computational Linguistics.