arXiv:2609.14542v1 Announce Type: new Abstract: Neyshekar is presented as an open Persian read-speech corpus designed for coverage of both formal and informal language, named entities, and longer utt...
PARSA‑Bench is the first dedicated benchmark for evaluating large audio‑language models on Persian, addressing unique challenges such as classical poetry, traditional music, and code‑switching. It comprises 16 tasks—10 of which are new—covering speech understanding, paralinguistic analysis, and culturally grounded audio reasoning. Across most tasks, text‑only baselines outperform audio‑based models, indicating that audio understanding remains the main limitation, except for Persian poetry where prosody provides additional information that audio beats text.
arXiv:2610.19388v1 Announce Type: new Abstract: Massively multilingual speech modeling has made remarkable progress in recent years, yet existing training resources and benchmarks still underrepresen...
Recent advances in zero-shot text-to-speech (TTS) have substantially improved speech quality and voice cloning fidelity. However, many zero-shot TTS systems still depend on audio prompt transcripts at inference time.
arXiv:2606. 24320v1 Announce Type: cross Abstract: We present ZONOS2 8B, our latest TTS model, which achieves state-of-the-art naturalness, prosody, and voice cloning fidelity.
GigaSpeech 3 is a new Arabic speech corpus comprising 39,000 hours and 10 million weakly transcribed segments across Algerian, Egyptian, Iraqi, Saudi, and Syrian dialects. It is built from 60 hours of manual transcription per dialect, using self‑supervised labeling models and noisy‑student refinement to produce high‑quality data. The corpus enables dialect‑specific and multi‑dialect speech recognition models that outperform existing third‑party systems, achieving a macro‑average word error rate of 23.86% for dialect‑specific models and 25.51% for a multi‑dialect model.
We present ZONOS2 8B, our latest TTS model, which achieves state-of-the-art naturalness, prosody, and voice cloning fidelity. We improve upon Zonos-v0.
The paper introduces a training‑free speech‑and‑text‑to‑pronunciation (ST2P) pipeline that combines lexical candidates from G2P tools with acoustic rescoring using frozen pretrained S2P models. By performing a left‑to‑right greedy search over whole‑sequence negative log‑likelihoods, the method achieves a dramatic reduction in character error rate on Japanese corpora, outperforming both baseline G2P/S2P approaches and commercial multimodal LLMs. The approach is also significantly faster—3–3.5× faster than beam search and twice as fast as direct decoding—while maintaining high accuracy across multiple languages.
arXiv:2606. 08843v1 Announce Type: cross Abstract: We present a voice conversion (VC) framework that utilizes K-Nearest Neighbors (KNN) retrieval over WavLM representations to align non-parallel source and target speech, constructing synthetic training pairs for supervised learning.
Lipi-Ghor is a large-scale, weakly-supervised Bengali speech corpus containing 885 hours of audio from 998 YouTube videos across 17 domains, each annotated in SSTT format with speaker diarization and timestamps. It is the first Bengali corpus to combine extensive coverage with speaker-diarized, long-form audio annotations, and the authors provide a replicable pipeline for building similar resources in other languages. The dataset includes a human-annotated gold evaluation split and is released under CC‑BY‑4.0, with raw audio withheld.
The paper introduces a unified phoneme‑based TTS‑to‑ASR augmentation pipeline that uses a multilingual TTS model with language‑ID conditioning and incorporates grapheme‑to‑phoneme conversion, reference‑speech filtering, and candidate‑text selection. It proposes phoneme‑frequency‑guided selection (PFGS) to rank sentences based on phoneme frequencies from real ASR labels, and demonstrates that random augmentation and PFGS both improve ASR performance across Arabic, French, Italian, and Portuguese test sets, with PFGS yielding up to a 19.3% relative WER reduction. The study also shows that filtering reference speech can further lower WER by up to 0.59 points on certain datasets.
arXiv:2507.02115v3 Announce Type: cross Abstract: Synthesizing second-language (L2) speech is potentially highly valued for L2 language learning experience and feedback. However, due to the lack of L...