Abstract:Full-duplex spoken dialogue systems can model natural conversational behaviours such as interruptions, overlaps, and backchannels, yet such systems remain largely unexplored for Indian languages. We present the first open, reproducible full-duplex spoken dialogue system for Hindi by adapting Moshi, a state-of-the-art duplex speech architecture, using a custom Hindi tokeniser and training on 26,000 hours of real spontaneous conversations collected from 14,695 speakers with separate speaker channels, enabling direct learning of turn-taking and overlap patterns from natural interactions. To support Hindi text generation, we replace the original English tokeniser and reinitialise text-vocabulary-dependent parameters while retaining the pre-trained audio components. We propose a two-stage training recipe -- large-scale pre-training followed by fine-tuning on 1,000 hours of conversational data. Evaluation through the prompted dialogue continuation paradigm with both automatic metrics and human judgments demonstrates that the resulting model generates natural and meaningful full-duplex conversational behaviour in Hindi. This work serves as a first step toward real-time duplex spoken dialogue systems for Hindi and other Indian languages.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2604.23295 [cs.CL] |
| (or arXiv:2604.23295v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2604.23295 arXiv-issued DOI via DataCite |
From: Bhaskar Singh [view email]
[v1]
Sat, 25 Apr 2026 13:18:40 UTC (249 KB)
[v2]
Mon, 25 May 2026 03:30:22 UTC (249 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。