Hallucination-Aware Multilingual Speech to-Speech Summarization using Reinforcement-Aligned Large Speech Language Models

Registration ID: IJNRD_325241

Published ID: IJNRD2605853

: https://doi.org/10.56975/ijnrd.v11i5.325241

About Hard Copy and Transparent Peer Review Report

S. Dinesh@Dhanabalan, S. Praveen Kumar, R. Ragupathy

Speech-to-Speech Multilingual Summarization; Speech Summarization; Large Speech Hallucination Language Models; Detection; Factual Consistency; Reinforcement Alignment; Direct Preference Optimization; Speaker Aware Summarization; Long-Form Speech Understanding; Multimodal Large Language Models

The rapid growth of long-form spoken content, including meetings, lectures, interviews, classroom discussions, and online consultations, has created a strong need for automatic systems that can generate concise, faithful, and accessible summaries from speech [1]. Existing speech summarization systems commonly depend on cascaded pipelines, where speech is first converted into text using automatic speech recognition and then summarized using text-based models. Although effective, these systems suffer from ASR error propagation, loss of speaker-level information, multilingual degradation, and text-only output limitations. In addition, large language model-based summarizers may generate hallucinated or unsupported information, which reduces the reliability of summaries in high-value domains such as education, healthcare, legal documentation, and business communication [2]. To address these issues, this paper proposes a hallucination-aware multilingual speech to-speech summarization framework using reinforcement-aligned large speech language models. The proposed system is designed to be evaluated using ROUGE, BERTScore, WER, CER, hallucination rate, factual consistency score, speaker attribution accuracy, multilingual adequacy, latency, and mean opinion score.Inspired by recent multilingual speech-to-speech modeling preference-alignment methods, and the framework applies Direct Preference Optimization or optimization to reinforcement-style improve factual consistency and reduce unsupported summary claims [3], [4]. The factuality verification module compares generated summaries with source speech/transcript evidence to identify contradictions, missing context, and speaker-attribution errors [5]. The proposed system is evaluated using ROUGE, BERTScore, WER, CER, hallucination rate, factual consistency score, speaker attribution accuracy, multilingual adequacy, latency, and mean opinion score. This study contributes a unified framework for faithful, multilingual, and speech-output oriented summarization of long-form spoken content.

Published Paper   E-Certificate

S. Dinesh@Dhanabalan, S. Praveen Kumar, & R. Ragupathy (May-2026). Hallucination-Aware Multilingual Speech to-Speech Summarization using Reinforcement-Aligned Large Speech Language Models . INTERNATIONAL JOURNAL OF NOVEL RESEARCH AND DEVELOPMENT, 11(5), i563-i583. https://doi.org/10.56975/ijnrd.v11i5.325241

Volume 11 Issue 5, May-2026

Pages : i563-i583

Paper Reg. ID: IJNRD_325241

Published Paper Id: IJNRD2605853

Research Area: Other area not in list

Author Type: Indian Author

Country: Thiruvarur, TamilNadu, India

Published Paper PDF: https://ijnrd.org/papers/IJNRD2605853.pdf

Published Paper URL: https://ijnrd.org/viewpaperforall?paper=IJNRD2605853

Crossref DOI: https://doi.org/10.56975/ijnrd.v11i5.325241

Journal Name: INTERNATIONAL JOURNAL OF NOVEL RESEARCH AND DEVELOPMENT(IJNRD)

UGC CARE JOURNAL PUBLICATION | ISSN: 2456-4184 | IMPACT FACTOR: 8.76 Calculated By Google Scholar | ESTD YEAR: 2016

An International UGC CARE JOURNAL PUBLICATION, Low Cost, Scholarly Open Access, Peer-Reviewed, Refereed Journal Impact Factor 8.76 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator

Publisher: IJNRD (IJ Publication) Janvi Wave | IJNRD.ORG | IJNRD.COM | IJPUB.ORG

© 2026 - Authors hold the copyright of this article. This work is licensed under a Creative Commons Attribution 4.0 International License. and The Open Definition. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (CC BY 4.0). 🛡️ Disclaimer: The content, data, and findings in this article are based on the authors’ research and have been peer-reviewed for academic purposes only. Readers are advised to verify all information before practical or commercial use. The journal and its editorial board are not liable for any errors, losses, or consequences arising from its use.

Call For Paper - Volume 11 | Issue 6 | June 2026

IJNRD is a Scholarly Open Access, Peer-Reviewed, Refereed, and UGC CARE Journal Publication with a High Impact Factor of 8.76 (calculated by Google Scholar & Semantic Scholar | AI-Powered Research Tool). It is a Multidisciplinary, Monthly, Low-Cost, and Transparent Peer Review Journal Publication that adheres to the New UGC CARE Transparent Peer-Reviewed Journal Policy and aligns with Scopus Journal Publication standards to ensure the highest level of research quality and credibility.

IJNRD offers comprehensive Journal Publication Services including indexing in all major databases and metadata repositories, Digital Object Identifier (Crossref DOI) assignment for each published article with additional fees, citation generation tools, and full Open Access visibility to enhance global research reach and citation impact.

The INTERNATIONAL JOURNAL OF NOVEL RESEARCH AND DEVELOPMENT (IJNRD) aims to advance applied, theoretical, and experimental research across diverse academic and professional fields. The journal promotes global knowledge exchange among researchers, developers, academicians, engineers, and practitioners, serving as a trusted platform for innovative, peer-reviewed journal publication and scientific collaboration.

Indexing Coverage: Google Scholar, SSRN, ResearcherID-Publons, Semantic Scholar (AI-Powered Research Tool), Microsoft Academic, Academia.edu, arXiv.org, ResearchGate, CiteSeerX, ResearcherID (Thomson Reuters), Mendeley, DocStoc, ISSUU, Scribd, and many other recognized academic repositories.

Transparent Peer Review Journal Publication: IJNRD operates a strict double-blind peer review system managed by 3000+ expert reviewers, ensuring ethical, unbiased, and high-quality review for every research paper.

For Indian Authors : Get a transparent peer review report from Scholar9.com for just ₹1000. View Sample Report

For Foreign Authors : A detailed peer review report is available through Scholar9.com for $20 USD. View Sample Report

Transparent Peer Review Journal Publication

⭐ Transparent Peer Review | 🕵️‍♂️ Double-Blind | 👨‍🏫 3000+ Expert Reviewers | 🇮🇳 Report for India Author ₹1000 | 🌐 Report for Foreign Author $20 | 📄 Sample Reports on Scholar9.com | 🌍 High Credibility | ⚖️ Ethical & Unbiased Evaluation

How to submit the paper?

By Our website

Click Here to Submit Paper Online

By Mail

Send your paper to editor@ijnrd.org

Recently, the UGC discontinued the UGC-CARE Journal List and introduced new parameters that allow publication in Transparent Peer-Reviewed (Refereed) Journals. IJNRD is Transparent Peer Review Journal Valid As per New UGC Notification.

You can now publish your research paper in IJNRD.ORG. IJNRD is a Transparent Peer-Reviewed Open Access (Refereed Journal), UGC and UGC CARE Approved, Crossref DOI, Multidisciplinary, Impact Factor calculate by Google Scholar. As an International, open-access, and online journal, Publishing with us ensures wider reach, academic credibility, and enhanced recognition for your work.

For more details, refer to the official notice: UGC Public Notice

⭐ Low Cost ₹1570 | 📚 UGC CARE Approved | 🔍 Peer-Reviewed | 🌐 Open Access | 🔗 Crossref DOI & Global Indexing | 📊 Google Scholar Impact Factor | 🧪 Multidisciplinary

Submit Paper Online  Call for Paper  About IJNRD UGC CARE Approval

Important Dates for Current issue

Paper Submission Open For: June 2026

Current Issue: Volume 11 | Issue 6 | June 2026

Impact Factor: 8.76

Last Date for Paper Submission: Till 30-Jun-2026

Notification of Review Result: Transparent peer review process - your paper is evaluated by experts, and you receive acceptance or rejection updates via email and SMS.

Publication of Paper: Once all documents are submitted, your paper is published without delay, and you can instantly download your certificate and confirmation letter online.

Frequency: Monthly (12 issue Annually).

Journal Type: IJNRD is an international open-access journal offering Low Cost Journal Publication, transparent Peer Review Journal Publication, Crossref DOI, and multidisciplinary research visibility under UGC CARE Approved Journal Publication.

Subject Category: Research Area

Approval, Licenses and Indexing: More Details

Call For Paper - Volume 11 | Issue 6 | June 2026

IJNRD.org offers low-cost journal publication starting at ₹1570 with UGC CARE Approved, refereed, peer-reviewed, open-access publishing. This multidisciplinary monthly journal, available in both online and print formats, features a strong Google Scholar-based impact factor of 8.76, Transparent Peer Review, CrossRef DOI, global indexing, fast publication, and complete metadata for maximum research visibility and citation impact across multidisciplinary domains.

Volume 11 | Issue 6 | June 2026 | IJNRD Transparent Peer Review Certificate | Submit Paper Online till 30-Jun-2026

⭐ UGC CARE Approved Refereed Journal | 🔍 Transparent Peer Review | 🌐 Open Access Publishing | 💰 Low-Cost ₹1570 | 🔗 CrossRef DOI & Global Indexing | 📊 Google Scholar Impact Factor 8.76 | 🧪 Multidisciplinary | Online & Print

Submit Paper Online