Clinical studies have shown that speech pauses can reflect the cognitive function differences between Alzheimer’s Disease (AD) and non-AD patients, while the value of pause information in AD detection has not been fully explored. Herein, we propose a speech pause feature extraction and encoding strategy for only acoustic-signal-based AD detection. First, a voice activity detection (VAD) method was constructed to detect pause/non-pause feature and encode it to binary pause sequences that are easier to calculate. Then, an ensemble machine-learning-based approach was proposed for the classification of AD from the participants’ spontaneous speech, based on the VAD Pause feature sequence and common acoustic feature sets (ComParE and eGeMAPS). The proposed pause feature sequence was verified in five machine-learning models. The validation data included two public challenge datasets (ADReSS and ADReSSo, English voice) and a local dataset (10 audio recordings containing five patients and five controls, Chinese voice). Results showed that the VAD Pause feature was more effective than common feature sets (ComParE: 6373 features and eGeMAPS: 88 features) for AD classification, and that the ensemble method improved the accuracy by more than 5% compared to several baseline methods (8% on the ADReSS dataset; 5.9% on the ADReSSo dataset). Moreover, the pause-sequence-based AD detection method could achieve 80% accuracy on the local dataset. Our study further demonstrated the potential of pause information in speech-based AD detection, and also contributed to a more accessible and general pause feature extraction and encoding method for AD detection.

Deep learning has significantly advanced text-to-speech (TTS) systems. These neural network-based systems have enhanced speech synthesis quality and are increasingly vital in applications like human-computer interaction. However, conventional TTS models still face challenges, as the synthesized speeches often lack naturalness and expressiveness. Additionally, the slow inference speed, reflecting low efficiency, contributes to the reduced voice quality. This paper introduces SynthRhythm-TTS (SR-TTS), an optimized Transformer-based structure designed to enhance synthesized speech. SR-TTS not only improves phonological quality and naturalness but also accelerates the speech generation process, thereby increasing inference efficiency. SR-TTS contains an encoder, a rhythm coordinator, and a decoder. In particular, a pre-duration predictor within the cadence coordinator and a self-attention-based feature predictor work together to enhance the naturalness and articulatory accuracy of speech. In addition, the introduction of causal convolution enhances the consistency of the time series. The cross-linguistic capability of SR-TTS is validated by training it on both English and Chinese corpora. Human evaluation shows that SR-TTS outperforms existing techniques in terms of speech quality and naturalness of expression. This technology is particularly suitable for applications that require high-quality natural speech, such as intelligent assistants, speech synthesized podcasts, and human-computer interaction.

Text-to-speech (TTS) synthesizers have been widely used as a vital assistive tool in various fields. Traditional sequence-to-sequence (seq2seq) TTS such as Tacotron2 uses a single soft attention mechanism for encoder and decoder alignment tasks, which is the biggest shortcoming that incorrectly or repeatedly generates words when dealing with long sentences. It may also generate sentences with run-on and wrong breaks regardless of punctuation marks, which causes the synthesized waveform to lack emotion and sound unnatural. In this paper, we propose an end-to-end neural generative TTS model that is based on the deep-inherited attention (DIA) mechanism along with an adjustable local-sensitive factor (LSF). The inheritance mechanism allows multiple iterations of the DIA by sharing the same training parameter, which tightens the token–frame correlation, as well as fastens the alignment process. In addition, LSF is adopted to enhance the context connection by expanding the DIA concentration region. In addition, a multi-RNN block is used in the decoder for better acoustic feature extraction and generation. Hidden-state information driven from the multi-RNN layers is utilized for attention alignment. The collaborative work of the DIA and multi-RNN layers contributes to outperformance in the high-quality prediction of the phrase breaks of the synthesized speech. We used WaveGlow as a vocoder for real-time, human-like audio synthesis. Human subjective experiments show that the DIA-TTS achieved a mean opinion score (MOS) of 4.48 in terms of naturalness. Ablation studies further prove the superiority of the DIA mechanism for the enhancement of phrase breaks and attention robustness.