arXiv:2607. 08111v1 Announce Type: cross Abstract: Training target speaker extraction (TSE) models for real conversational mixtures remains challenging because large-scale training corpora and clean target speech for supervision are unavailable.
arXiv:2606. 03957v1 Announce Type: cross Abstract: Conversational ASR for lower-resource languages and niche domains is limited by the scarcity of domain-matched multi-speaker training data.
Conversational ASR for lower-resource languages and niche domains is limited by the scarcity of domain-matched multi-speaker training data. We propose an augmentation pipeline that generates scenario-level dialogues with participant metadata, maps speaker attributes to TTS voice profiles, and assembles synthesized utterances into speaker-aware simulated conversations.
arXiv:2606. 19325v1 Announce Type: cross Abstract: Existing multi-speaker dialogue systems bind speakers to utterances through structured supervision: per-turn tags, multi-stream transcriptions, or learnable speaker embeddings.
arXiv:2512. 20978v2 Announce Type: replace-cross Abstract: Language Model (LM)-based generative modeling has emerged as a promising direction for TSE, offering potential for improved generalization and high-fidelity speech.
Existing multi-speaker dialogue systems bind speakers to utterances through structured supervision: per-turn tags, multi-stream transcriptions, or learnable speaker embeddings. These systems operate within speech-only pipelines that produce clean vocal sequences without the ambient texture of real conversations.
The paper introduces a method called randomized intermediate guidance for training tandem speech-to-speech models, where a large language model (LLM) acts as a backend providing candidate responses while the user is speaking. Instead of simulating the backend’s guidance, the approach derives guidance directly from the conversation corpus, using target responses for informative guidance and randomly sampled responses to simulate irrelevant updates. Experiments on synthetic dialogues and 3.8k hours of real conversations show that this technique yields response quality comparable to LLM-generated baselines while improving natural turn‑taking and audio‑judge naturalness.
The paper addresses challenges in extracting target and multiple speakers from real conversational speech, noting that real conversations contain more silence and enrolment samples that differ from the target speech. It introduces a new loss function that reduces the impact of excess silence during training, yielding improvements in STOI (from 0.55 to 0.60) and frequency‑weighted segmental SNR (from 4.35 to 5.12). The study also investigates how mismatches between enrolment and target speech affect performance.
The paper introduces asymmetric classifier‑free guidance (CFG) for target‑speaker ASR using Whisper, where a speaker‑conditioned branch predicts the target transcript and a speaker‑unconditioned branch predicts serialized multi‑speaker transcripts. CFG modulates the influence of speaker conditioning during decoding via a single guidance scale, which is first set globally on development data and then refined per utterance by a lightweight encoder‑based predictor while keeping the recognition model fixed. The resulting system yields up to 21.8% relative WER reduction over a condition‑only baseline and 5.6% over standard conditional decoding under domain shifts.
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.
X-VC is a zero‑shot streaming voice conversion system that performs one‑step conversion directly in the latent space of a pretrained neural codec. It employs a dual‑conditioning acoustic converter that jointly models source codec latents and target acoustic conditions, while using adaptive normalization to inject utterance‑level speaker information. The model is trained with generated paired data and a role‑assignment strategy, and uses a chunkwise inference scheme with overlap smoothing to achieve low‑latency streaming inference, achieving superior WER, speaker similarity, and real‑time factor on the Seed‑TTS‑Eval benchmark.
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.
arXiv:2604. 19635v2 Announce Type: replace-cross Abstract: While generative models have set new benchmarks for Target Speaker Extraction (TSE), their inherent reliance on global context precludes deployment in real-time applications.