Hybrid language models that combine Attention and State Space Models (SSMs) have been shown to achieve state-of-the-art accuracy and runtime performance. Recent work has also demonstrated that applying pruning and distillation to Attention-only models yields smaller, more accurate models at a fraction of the training cost. In this work, we explore the effectiveness of compressing Hybrid architectures. To this end, we introduce a novel group-aware pruning method for Mamba layers that preserves the structural integrity of SSM blocks and their sequence modeling capabilities. We combine this method with FFN, embedding dimension, and layer pruning, along with knowledge distillation-based retraining to obtain a unified compression recipe for hybrid models. Using this recipe, we compress the Nemotron-H 8B Hybrid model down to 4B parameters with up to $40\times$ fewer training tokens compared to similarly-sized models. The resulting model surpasses the accuracy of similarly-sized models while achieving $\sim2\times$ faster inference throughput, significantly advancing the Pareto frontier.

Recently normalizing flows have been gaining traction in text-to-speech (TTS) and voice conversion (VC) due to their state-of-the-art (SOTA) performance. Normalizing flows are unsupervised generative models. In this paper, we introduce supervision to the training process of normalizing flows, without the need for parallel data. We call this training paradigm AutoEncoder Normalizing Flow (AE-Flow). It adds a reconstruction loss forcing the model to use information from the conditioning to reconstruct an audio sample. Our goal is to understand the impact of each component and find the right combination of the negative log-likelihood (NLL) and the reconstruction loss in training normalizing flows with coupling blocks. For that reason we will compare flow-based mapping model trained with: (i) NLL loss, (ii) NLL and reconstruction losses, as well as (iii) reconstruction loss only. Additionally, we compare our model with SOTA VC baseline. The models are evaluated in terms of naturalness, speaker similarity, intelligibility in many-to-many and many-to-any VC settings. The results show that the proposed training paradigm systematically improves speaker similarity and naturalness when compared to regular training methods of normalizing flows. Furthermore, we show that our method improves speaker similarity and intelligibility over the state-of-the-art.

Non-parallel voice conversion (VC) is typically achieved using lossy representations of the source speech. However, ensuring only speaker identity information is dropped whilst all other information from the source speech is retained is a large challenge. This is particularly challenging in the scenario where at inference-time we have no knowledge of the text being read, i.e., text-free VC. To mitigate this, we investigate information-preserving VC approaches.Normalising flows have gained attention for text-to-speech synthesis, however have been under-explored for VC. Flows utilize invertible functions to learn the likelihood of the data, thus provide a lossless encoding of speech. We investigate normalising flows for VC in both text-conditioned and text-free scenarios. Furthermore, for text-free VC we compare pre-trained and jointly-learnt priors. Flow-based VC evaluations show no degradation between text-free and text-conditioned VC, resulting in improvements over the state-of-the-art. Also, joint-training of the prior is found to negatively impact text-free VC quality.

Creating realistic and natural-sounding synthetic speech remains a big challenge for voice identities unseen during training. As there is growing interest in synthesizing voices of new speakers, here we investigate the ability of normalizing flows in text-to-speech (TTS) and voice conversion (VC) modes to extrapolate from speakers observed during training to create unseen speaker identities. Firstly, we create an approach for TTS and VC, and then we comprehensively evaluate our methods and baselines in terms of intelligibility, naturalness, speaker similarity, and ability to create new voices. We use both objective and subjective metrics to benchmark our techniques on 2 evaluation tasks: zero-shot and new voice speech synthesis. The goal of the former task is to measure the precision of the conversion to an unseen voice. The goal of the latter is to measure the ability to create new voices. Extensive evaluations demonstrate that the proposed approach systematically allows to obtain state-of-the-art performance in zero-shot speech synthesis and creates various new voices, unobserved in the training set. We consider this work to be the first attempt to synthesize new voices based on mel-spectrograms and normalizing flows, along with a comprehensive analysis and comparison of the TTS and VC modes.

In expressive speech synthesis it is widely adopted to use latent prosody representations to deal with variability of the data during training. Same text may correspond to various acoustic realizations, which is known as a one-to-many mapping problem in text-to-speech. Utterance, word, or phoneme-level representations are extracted from target signal in an auto-encoding setup, to complement phonetic input and simplify that mapping. This paper compares prosodic embeddings at different levels of granularity and examines their prediction from text. We show that utterance-level embeddings have insufficient capacity and phoneme-level tend to introduce instabilities when predicted from text. Word-level representations impose balance between capacity and predictability. As a result, we close the gap in naturalness by 90% between synthetic speech and recordings on LibriTTS dataset, without sacrificing intelligibility.

Regional accents of the same language affect not only how words are pronounced (i.e., phonetic content), but also impact prosodic aspects of speech such as speaking rate and intonation. This paper investigates a novel flow-based approach to accent conversion using normalizing flows. The proposed approach revolves around three steps: remapping the phonetic conditioning, to better match the target accent, warping the duration of the converted speech, to better suit the target phonemes, and an attention mechanism that implicitly aligns source and target speech sequences. The proposed remap-warp-attend system enables adaptation of both phonetic and prosodic aspects of speech while allowing for source and converted speech signals to be of different lengths. Objective and subjective evaluations show that the proposed approach significantly outperforms a competitive CopyCat baseline model in terms of similarity to the target accent, naturalness and intelligibility.

The research community has long studied computer-assisted pronunciation training (CAPT) methods in non-native speech. Researchers focused on studying various model architectures, such as Bayesian networks and deep learning methods, as well as on the analysis of different representations of the speech signal. Despite significant progress in recent years, existing CAPT methods are not able to detect pronunciation errors with high accuracy (only 60% precision at 40%–80% recall). One of the key problems is the low availability of mispronounced speech that is needed for the reliable training of pronunciation error detection models. If we had a generative model that could mimic non-native speech and produce any amount of training data, then the task of detecting pronunciation errors would be much easier. We present three innovative techniques based on phoneme-to-phoneme (P2P), text-to-speech (T2S) and speech-to-speech (S2S) conversion to generate correctly pronounced and mispronounced synthetic speech. We show that these techniques not only improve the accuracy of three machine learning models for detecting pronunciation errors, but also help establish a new state-of-the-art in the field. Earlier studies have used simple speech generation techniques such as P2P conversion, but only as an additional mechanism to improve the accuracy of pronunciation error detection. We, on the other hand, consider speech generation to be the first-class method of detecting pronunciation errors. The effectiveness of these techniques is assessed in the tasks of detecting pronunciation and lexical stress errors. Non-native English speech corpora of German, Italian, and Polish speakers are used in the evaluations. The best proposed S2S technique improves the accuracy of detecting pronunciation errors in AUC metric by 41% from 0.528 to 0.749 compared to the state-of-the-art approach.

We present a universal neural vocoder based on Parallel WaveNet, with an additional conditioning network called Audio Encoder. Our universal vocoder offers real-time high-quality speech synthesis on a wide range of use cases. We tested it on 43 internal speakers of diverse age and gender, speaking 20 languages in 17 unique styles, of which 7 voices and 5 styles were not exposed during training. We show that the proposed universal vocoder significantly outperforms speaker-dependent vocoders overall. We also show that the proposed vocoder outperforms several existing neural vocoder architectures in terms of naturalness and universality. These findings are consistent when we further test on more than 300 open-source voices.

A common approach to the automatic detection of mispronunciation in language learning is to recognize the phonemes produced by a student and compare it to the expected pronunciation of a native speaker. This approach makes two simplifying assumptions: a) phonemes can be recognized from speech with high accuracy, b) there is a single correct way for a sentence to be pronounced. These assumptions do not always hold, which can result in a significant amount of false mispronunciation alarms. We propose a novel approach to overcome this problem based on two principles: a) taking into account uncertainty in the automatic phoneme recognition step, b) accounting for the fact that there may be multiple valid pronunciations. We evaluate the model on non-native (L2) English speech of German, Italian and Polish speakers, where it is shown to increase the precision of detecting mispronunciations by up to 18% (relative) compared to the common approach.

This paper proposes a general enhancement to the Normalizing Flows (NF) used in neural vocoding. As a case study, we improve expressive speech vocoding with a revamped Parallel Wavenet (PW). Specifically, we propose to extend the affine transformation of PW to the more expressive invertible non-affine function. The greater expressiveness of the improved PW leads to better-perceived signal quality and naturalness in the waveform reconstruction and text-to-speech (TTS) tasks. We evaluate the model across different speaking styles on a multi-speaker, multi-lingual dataset. In the waveform reconstruction task, the proposed model closes the naturalness and signal quality gap from the original PW to recordings by $10\%$, and from other state-of-the-art neural vocoding systems by more than $60\%$. We also demonstrate improvements in objective metrics on the evaluation test set with L2 Spectral Distance and Cross-Entropy reduced by $3\%$ and $6\unicode{x2030}$ comparing to the affine PW. Furthermore, we extend the probability density distillation procedure proposed by the original PW paper, so that it works with any non-affine invertible and differentiable function.

This paper describes two novel complementary techniques that improve the detection of lexical stress errors in non-native (L2) English speech: attention-based feature extraction and data augmentation based on Neural Text-To-Speech (TTS). In a classical approach, audio features are usually extracted from fixed regions of speech such as the syllable nucleus. We propose an attention-based deep learning model that automatically derives optimal syllable-level representation from frame-level and phoneme-level audio features. Training this model is challenging because of the limited amount of incorrect stress patterns. To solve this problem, we propose to augment the training set with incorrectly stressed words generated with Neural TTS. Combining both techniques achieves 94.8% precision and 49.2% recall for the detection of incorrectly stressed words in L2 English speech of Slavic and Baltic speakers.

We propose a weakly-supervised model for word-level mispronunciation detection in non-native (L2) English speech. To train this model, phonetically transcribed L2 speech is not required and we only need to mark mispronounced words. The lack of phonetic transcriptions for L2 speech means that the model has to learn only from a weak signal of word-level mispronunciations. Because of that and due to the limited amount of mispronounced L2 speech, the model is more likely to overfit. To limit this risk, we train it in a multi-task setup. In the first task, we estimate the probabilities of word-level mispronunciation. For the second task, we use a phoneme recognizer trained on phonetically transcribed L1 speech that is easily accessible and can be automatically annotated. Compared to state-of-the-art approaches, we improve the accuracy of detecting word-level pronunciation errors in AUC metric by 30% on the GUT Isle Corpus of L2 Polish speakers, and by 21.5% on the Isle Corpus of L2 German and Italian speakers.

This paper proposed a novel approach for the detection and reconstruction of dysarthric speech. The encoder-decoder model factorizes speech into a low-dimensional latent space and encoding of the input text. We showed that the latent space conveys interpretable characteristics of dysarthria, such as intelligibility and fluency of speech. MUSHRA perceptual test demonstrated that the adaptation of the latent space let the model generate speech of improved fluency. The multi-task supervised approach for predicting both the probability of dysarthric speech and the mel-spectrogram helps improve the detection of dysarthria with higher accuracy. This is thanks to a low-dimensional latent space of the auto-encoder as opposed to directly predicting dysarthria from a highly dimensional mel-spectrogram.

Statistical TTS systems that directly predict the speech waveform have recently reported improvements in synthesis quality. This investigation evaluates Amazon's statistical speech waveform synthesis (SSWS) system. An in-depth evaluation of SSWS is conducted across a number of domains to better understand the consistency in quality. The results of this evaluation are validated by repeating the procedure on a separate group of testers. Finally, an analysis of the nature of speech errors of SSWS compared to hybrid unit selection synthesis is conducted to identify the strengths and weaknesses of SSWS. Having a deeper insight into SSWS allows us to better define the focus of future work to improve this new technology.