The paper adapts Reinforce Adjoint Matching (RAM) to generative speech enhancement, allowing a pretrained model to be post‑trained on real recordings using weak supervision such as text transcripts. RAM shifts the model’s conditional distribution toward higher‑reward outputs by generating enhanced speech on‑policy, evaluating each output with a potentially non‑differentiable reward, and analytically re‑noising the endpoint to create inputs for a reward‑guided regression objective. Experiments on real CHiME‑4 recordings show a 5.08‑percentage‑point reduction in word error rate compared to the pretrained FlowSE model, while maintaining all reported non‑intrusive speech quality metrics and receiving no significant preference in a subjective listening test.
arXiv:2510. 20441v2 Announce Type: replace-cross Abstract: Neural audio codecs have largely promoted the application of language models (LMs) for speech applications.
arXiv:2603. 13952v3 Announce Type: replace-cross Abstract: In existing Audio-Visual Speech Enhancement (AVSE) methods, objectives such as Scale-Invariant Signal-to-Noise Ratio (SI-SNR) and Mean Squared Error (MSE) are widely used; however, their correlation with perceived speech quality is often suboptimal and provides limited interpretability for optimization.
arXiv:2606. 07080v1 Announce Type: cross Abstract: We present dots.
arXiv:2608.31035v1 Announce Type: new Abstract: Codec-based text-to-speech (TTS) models make language-model post-training applicable to speech generation, but it remains unclear when learned perceptu...
The study examines how speech preprocessing—such as enhancement, sample selection, and demographic balancing—affects Alzheimer’s disease detection models that use the Pitt Corpus. Experiments reveal that while speech‑enhanced datasets boost in‑domain accuracy, they diminish cross‑dataset robustness and introduce class imbalance and prediction shifts, even when training and testing enhancements are matched. Large audio‑language models show similar sensitivity, indicating that cleaner speech does not guarantee better real‑world performance.
The paper introduces Corrective Forcing (CoF), a post‑training method that aligns diffusion and flow generative models for speech enhancement by training them on self‑generated rollout states. CoF corrects predictions toward ground truth under dynamic sampling schedules and regularizes local evolution with counterfactual transitions, applying a unified objective across both model types. Experiments on SB‑VE and OT‑CFM show improved perceptual quality, reconstruction fidelity, and robustness to varying sampling steps.
arXiv:2511. 11686v4 Announce Type: replace Abstract: Speech enhancement (SE) requires high-fidelity reconstruction of clean speech that preserves linguistic and paralinguistic cues while maintaining high perceptual quality.
Speech-based Alzheimer's disease (AD) detection increasingly relies on speech-enhanced and curated versions of the Pitt Corpus, where speech enhancement, sample selection, and demographic balancing ar...
The paper proposes a single‑utterance test‑time adaptation (TTA) method for speech enhancement that uses an autoregressive prior trained on clean speech latent representations from a neural audio codec. The adaptation regularizes a pretrained enhancement model by minimizing the Kullback‑Leibler divergence between the enhanced speech distribution and the clean speech prior. Experiments on multiple noisy speech datasets demonstrate consistent improvements in speech quality, especially when training and testing noise conditions differ.
arXiv:2609.13150v1 Announce Type: cross Abstract: Reference-free quality predictors such as UTMOS, DNSMOS and SCOREQ are the de facto automatic evaluators for text-to-speech (TTS) and are increasingl...
arXiv:2609.26536v1 Announce Type: new Abstract: In LLM-based speech translation, transcription-based chain-of-thought (CoT) suffers from a mismatch between reference transcripts used in supervised fi...
arXiv:2606. 09234v1 Announce Type: cross Abstract: Recent state-of-the-art (SOTA) text-to-speech (TTS) systems typically adopt a cascaded pipeline consisting of a speech tokenizer, an autoregressive large language model (LLM), and a diffusion based flow-matching (FM) model, with these components trained independently.