RALL-E: Robust Codec Language Modeling with Chain-of-Thought Prompting for Text-to-Speech Synthesis

We present RALL-E, a robust language modeling method for text-to-speech (TTS) synthesis. While previous work based on large language models (LLMs) shows impressive performance on zero-shot TTS, such methods often suffer from poor robustness, such as unstable prosody (weird pitch and rhythm/duration) and a high word error rate (WER), due to the autoregressive prediction style of language models. The core idea behind RALL-E is chain-of-thought (CoT) prompting, which decomposes the task into simpler steps to enhance the robustness of LLM-based TTS. To accomplish this idea, RALL-E first predicts prosody features (pitch and duration) of the input text and uses them as intermediate conditions to predict speech tokens in a CoT style. Second, RALL-E utilizes the predicted duration prompt to guide the computing of self-attention weights in Transformer to enforce the model to focus on the corresponding phonemes and prosody features when predicting speech tokens. Results of comprehensive objective and subjective evaluations demonstrate that, compared to a powerful baseline method VALL-E, RALL-E significantly improves the WER of zero-shot TTS from $6.3\%$ (without reranking) and $2.1\%$ (with reranking) to $2.8\%$ and $1.0\%$, respectively. Furthermore, we demonstrate that RALL-E correctly synthesizes sentences that are hard for VALL-E and reduces the error rate from $68\%$ to $4\%$.

Building speech corpus with diverse voice characteristics for its prompt-based representation

In text-to-speech synthesis, the ability to control voice characteristics is vital for various applications. By leveraging thriving text prompt-based generation techniques, it should be possible to enhance the nuanced control of voice characteristics. While previous research has explored the prompt-based manipulation of voice characteristics, most studies have used pre-recorded speech, which limits the diversity of voice characteristics available. Thus, we aim to address this gap by creating a novel corpus and developing a model for prompt-based manipulation of voice characteristics in text-to-speech synthesis, facilitating a broader range of voice characteristics. Specifically, we propose a method to build a sizable corpus pairing voice characteristics descriptions with corresponding speech samples. This involves automatically gathering voice-related speech data from the Internet, ensuring its quality, and manually annotating it using crowdsourcing. We implement this method with Japanese language data and analyze the results to validate its effectiveness. Subsequently, we propose a construction method of the model to retrieve speech from voice characteristics descriptions based on a contrastive learning method. We train the model using not only conservative contrastive learning but also feature prediction learning to predict quantitative speech features corresponding to voice characteristics. We evaluate the model performance via experiments with the corpus we constructed above.

Real-time Speech Extraction Using Spatially Regularized Independent Low-rank Matrix Analysis and Rank-constrained Spatial Covariance Matrix Estimation

Real-time speech extraction is an important challenge with various applications such as speech recognition in a human-like avatar/robot. In this paper, we propose the real-time extension of a speech extraction method based on independent low-rank matrix analysis (ILRMA) and rank-constrained spatial covariance matrix estimation (RCSCME). The RCSCME-based method is a multichannel blind speech extraction method that demonstrates superior speech extraction performance in diffuse noise environments. To improve the performance, we introduce spatial regularization into the ILRMA part of the RCSCME-based speech extraction and design two regularizers. Speech extraction experiments demonstrated that the proposed methods can function in real time and the designed regularizers improve the speech extraction performance.

SpeechBERTScore: Reference-Aware Automatic Evaluation of Speech Generation Leveraging NLP Evaluation Metrics

While subjective assessments have been the gold standard for evaluating speech generation, there is a growing need for objective metrics that are highly correlated with human subjective judgments due to their cost efficiency. This paper proposes reference-aware automatic evaluation methods for speech generation inspired by evaluation metrics in natural language processing. The proposed SpeechBERTScore computes the BERTScore for self-supervised dense speech features of the generated and reference speech, which can have different sequential lengths. We also propose SpeechBLEU and SpeechTokenDistance, which are computed on speech discrete tokens. The evaluations on synthesized speech show that our method correlates better with human subjective ratings than mel cepstral distortion and a recent mean opinion score prediction model. Also, they are effective in noisy speech evaluation and have cross-lingual applicability.

Localizing Acoustic Energy in Sound Field Synthesis by Directionally Weighted Exterior Radiation Suppression

A method for synthesizing the desired sound field while suppressing the exterior radiation power with directional weighting is proposed. The exterior radiation from the loudspeakers in sound field synthesis systems can be problematic in practical situations. Although several methods to suppress the exterior radiation have been proposed, suppression in all outward directions is generally difficult, especially when the number of loudspeakers is not sufficiently large. We propose the directionally weighted exterior radiation representation to prioritize the suppression directions by incorporating it into the optimization problem of sound field synthesis. By using the proposed representation, the exterior radiation in the prioritized directions can be significantly reduced while maintaining high interior synthesis accuracy, owing to the relaxed constraint on the exterior radiation. Its performance is evaluated with the application of the proposed representation to amplitude matching in numerical experiments.

JVNV: A Corpus of Japanese Emotional Speech with Verbal Content and Nonverbal Expressions

We present the JVNV, a Japanese emotional speech corpus with verbal content and nonverbal vocalizations whose scripts are generated by a large-scale language model. Existing emotional speech corpora lack not only proper emotional scripts but also nonverbal vocalizations (NVs) that are essential expressions in spoken language to express emotions. We propose an automatic script generation method to produce emotional scripts by providing seed words with sentiment polarity and phrases of nonverbal vocalizations to ChatGPT using prompt engineering. We select 514 scripts with balanced phoneme coverage from the generated candidate scripts with the assistance of emotion confidence scores and language fluency scores. We demonstrate the effectiveness of JVNV by showing that JVNV has better phoneme coverage and emotion recognizability than previous Japanese emotional speech corpora. We then benchmark JVNV on emotional text-to-speech synthesis using discrete codes to represent NVs. We show that there still exists a gap between the performance of synthesizing read-aloud speech and emotional speech, and adding NVs in the speech makes the task even harder, which brings new challenges for this task and makes JVNV a valuable resource for relevant works in the future. To our best knowledge, JVNV is the first speech corpus that generates scripts automatically using large language models.

Coco-Nut: Corpus of Japanese Utterance and Voice Characteristics Description for Prompt-based Control

In text-to-speech, controlling voice characteristics is important in achieving various-purpose speech synthesis. Considering the success of text-conditioned generation, such as text-to-image, free-form text instruction should be useful for intuitive and complicated control of voice characteristics. A sufficiently large corpus of high-quality and diverse voice samples with corresponding free-form descriptions can advance such control research. However, neither an open corpus nor a scalable method is currently available. To this end, we develop Coco-Nut, a new corpus including diverse Japanese utterances, along with text transcriptions and free-form voice characteristics descriptions. Our methodology to construct this corpus consists of 1) automatic collection of voice-related audio data from the Internet, 2) quality assurance, and 3) manual annotation using crowdsourcing. Additionally, we benchmark our corpus on the prompt embedding model trained by contrastive speech-text learning.

Do learned speech symbols follow Zipf's law?

In this study, we investigate whether speech symbols, learned through deep learning, follow Zipf's law, akin to natural language symbols. Zipf's law is an empirical law that delineates the frequency distribution of words, forming fundamentals for statistical analysis in natural language processing. Natural language symbols, which are invented by humans to symbolize speech content, are recognized to comply with this law. On the other hand, recent breakthroughs in spoken language processing have given rise to the development of learned speech symbols; these are data-driven symbolizations of speech content. Our objective is to ascertain whether these data-driven speech symbols follow Zipf's law, as the same as natural language symbols. Through our investigation, we aim to forge new ways for the statistical analysis of spoken language processing.

Diversity-based core-set selection for text-to-speech with linguistic and acoustic features

This paper proposes a method for extracting a lightweight subset from a text-to-speech (TTS) corpus ensuring synthetic speech quality. In recent years, methods have been proposed for constructing large-scale TTS corpora by collecting diverse data from massive sources such as audiobooks and YouTube. Although these methods have gained significant attention for enhancing the expressive capabilities of TTS systems, they often prioritize collecting vast amounts of data without considering practical constraints like storage capacity and computation time in training, which limits the available data quantity. Consequently, the need arises to efficiently collect data within these volume constraints. To address this, we propose a method for selecting the core subset~(known as \textit{core-set}) from a TTS corpus on the basis of a \textit{diversity metric}, which measures the degree to which a subset encompasses a wide range. Experimental results demonstrate that our proposed method performs significantly better than the baseline phoneme-balanced data selection across language and corpus size.

Kernel Interpolation of Incident Sound Field in Region Including Scattering Objects

A method for estimating the incident sound field inside a region containing scattering objects is proposed. The sound field estimation method has various applications, such as spatial audio capturing and spatial active noise control; however, most existing methods do not take into account the presence of scatterers within the target estimation region. Although several techniques exist that employ knowledge or measurements of the properties of the scattering objects, it is usually difficult to obtain them precisely in advance, and their properties may change during the estimation process. Our proposed method is based on the kernel ridge regression of the incident field, with a separation from the scattering field represented by a spherical wave function expansion, thus eliminating the need for prior modeling or measurements of the scatterers. Moreover, we introduce a weighting matrix to induce smoothness of the scattering field in the angular direction, which alleviates the effect of the truncation order of the expansion coefficients on the estimation accuracy. Experimental results indicate that the proposed method achieves a higher level of estimation accuracy than the kernel ridge regression without separation.