Electrical Engineering And Systems Science

Informed source separation has recently gained renewed interest with the introduction of neural networks and the availability of large multitrack datasets containing both the mixture and the separated sources. These approaches use prior information about the target source to improve separation. Historically, Music Information Retrieval researchers have focused primarily on score-informed source separation, but more recent approaches explore lyrics-informed source separation. However, because of the lack of multitrack datasets with time-aligned lyrics, models use weak conditioning with non-aligned lyrics. In this paper, we present a multimodal multitrack dataset with lyrics aligned in time at the word level with phonetic information as well as explore strong conditioning using the aligned phonemes. Our model follows a U-Net architecture and takes as input both the magnitude spectrogram of a musical mixture and a matrix with aligned phonetic information. The phoneme matrix is embedded to obtain the parameters that control Feature-wise Linear Modulation (FiLM) layers. These layers condition the U-Net feature maps to adapt the separation process to the presence of different phonemes via affine transformations. We show that phoneme conditioning can be successfully applied to improve singing voice source separation.

Text-based speech editors expedite the process of editing speech recordings by permitting editing via intuitive cut, copy, and paste operations on a speech transcript. A major drawback of current systems, however, is that edited recordings often sound unnatural because of prosody mismatches around edited regions. In our work, we propose a new context-aware method for more natural sounding text-based editing of speech. To do so, we 1) use a series of neural networks to generate salient prosody features that are dependent on the prosody of speech surrounding the edit and amenable to fine-grained user control 2) use the generated features to control a standard pitch-shift and time-stretch method and 3) apply a denoising neural network to remove artifacts induced by the signal manipulation to yield a high-fidelity result. We evaluate our approach using a subjective listening test, provide a detailed comparative analysis, and conclude several interesting insights.

Mispronunciation detection is an essential component of the Computer-Assisted Pronunciation Training (CAPT) systems. State-of-the-art mispronunciation detection models use Deep Neural Networks (DNN) for acoustic modeling, and a Goodness of Pronunciation (GOP) based algorithm for pronunciation scoring. However, GOP based scoring models have two major limitations: i.e., (i) They depend on forced alignment which splits the speech into phonetic segments and independently use them for scoring, which neglects the transitions between phonemes within the segment; (ii) They only focus on phonetic segments, which fails to consider the context effects across phonemes (such as liaison, omission, incomplete plosive sound, etc.). In this work, we propose the Context-aware Goodness of Pronunciation (CaGOP) scoring model. Particularly, two factors namely the transition factor and the duration factor are injected into CaGOP scoring. The transition factor identifies the transitions between phonemes and applies them to weight the frame-wise GOP. Moreover, a self-attention based phonetic duration modeling is proposed to introduce the duration factor into the scoring model. The proposed scoring model significantly outperforms baselines, achieving 20% and 12% relative improvement over the GOP model on the phoneme-level and sentence-level mispronunciation detection respectively.

Continuous speech separation plays a vital role in complicated speech related tasks such as conversation transcription. The separation model extracts a single speaker signal from a mixed speech. In this paper, we use transformer and conformer in lieu of recurrent neural networks in the separation system, as we believe capturing global information with the self-attention based method is crucial for the speech separation. Evaluating on the LibriCSS dataset, the conformer separation model achieves state of the art results, with a relative 23.5% word error rate (WER) reduction from bi-directional LSTM (BLSTM) in the utterance-wise evaluation and a 15.4% WER reduction in the continuous evaluation.

Speech synthesis has recently seen significant improvements in fidelity, driven by the advent of neural vocoders and neural prosody generators. However, these systems lack intuitive user controls over prosody, making them unable to rectify prosody errors (e.g., misplaced emphases and contextually inappropriate emotions) or generate prosodies with diverse speaker excitement levels and emotions. We address these limitations with a user-controllable, context-aware neural prosody generator. Given a real or synthesized speech recording, our model allows a user to input prosody constraints for certain time frames and generates the remaining time frames from input text and contextual prosody. We also propose a pitch-shifting neural vocoder to modify input speech to match the synthesized prosody. Through objective and subjective evaluations we show that we can successfully incorporate user control into our prosody generation model without sacrificing the overall naturalness of the synthesized speech.

Modern neural text-to-speech (TTS) synthesis can generate speech that is indistinguishable from natural speech. However, the prosody of generated utterances often represents the average prosodic style of the database instead of having wide prosodic variation. Moreover, the generated prosody is solely defined by the input text, which does not allow for different styles for the same sentence. In this work, we train a sequence-to-sequence neural network conditioned on acoustic speech features to learn a latent prosody space with intuitive and meaningful dimensions. Experiments show that a model conditioned on sentence-wise pitch, pitch range, phone duration, energy, and spectral tilt can effectively control each prosodic dimension and generate a wide variety of speaking styles, while maintaining similar mean opinion score (4.23) to our Tacotron baseline (4.26).

Transformer has achieved competitive performance against state-of-the-art end-to-end models in automatic speech recognition (ASR), and requires significantly less training time than RNN-based models. The original Transformer, with encoder-decoder architecture, is only suitable for offline ASR. It relies on an attention mechanism to learn alignments, and encodes input audio bidirectionally. The high computation cost of Transformer decoding also limits its use in production streaming systems. To make Transformer suitable for streaming ASR, we explore Transducer framework as a streamable way to learn alignments. For audio encoding, we apply unidirectional Transformer with interleaved convolution layers. The interleaved convolution layers are used for modeling future context which is important to performance. To reduce computation cost, we gradually downsample acoustic input, also with the interleaved convolution layers. Moreover, we limit the length of history context in self-attention to maintain constant computation cost for each decoding step. We show that this architecture, named Conv-Transformer Transducer, achieves competitive performance on LibriSpeech dataset (3.6\% WER on test-clean) without external language models. The performance is comparable to previously published streamable Transformer Transducer and strong hybrid streaming ASR systems, and is achieved with smaller look-ahead window (140~ms), fewer parameters and lower frame rate.

The effects of adding pitch and voice quality features such as jitter and shimmer to a state-of-the-art CNN model for Automatic Speech Recognition are studied in this work. Pitch features have been previously used for improving classical HMM and DNN baselines, while jitter and shimmer parameters have proven to be useful for tasks like speaker or emotion recognition. Up to our knowledge, this is the first work combining such pitch and voice quality features with modern convolutional architectures, showing improvements up to 7% and 3% relative WER points, for the publicly available Spanish Common Voice and LibriSpeech 100h datasets, respectively. Particularly, our work combines these features with mel-frequency spectral coefficients (MFSCs) to train a convolutional architecture with Gated Linear Units (Conv GLUs). Such models have shown to yield small word error rates, while being very suitable for parallel processing for online streaming recognition use cases. We have added pitch and voice quality functionality to Facebook's wav2letter speech recognition framework, and we provide with such code and recipes to the community, to carry on with further experiments. Besides, to the best of our knowledge, our Spanish Common Voice recipe is the first public Spanish recipe for wav2letter.

In this paper, we propose a semi-supervised learning (SSL) technique for training deep neural networks (DNNs) to generate speaker-discriminative acoustic embeddings (speaker embeddings). Obtaining large amounts of speaker recognition train-ing data can be difficult for desired target domains, especially under privacy constraints. The proposed technique reduces requirements for labelled data by leveraging unlabelled data. The technique is a variant of virtual adversarial training (VAT) [1] in the form of a loss that is defined as the robustness of the speaker embedding against input perturbations, as measured by the cosine-distance. Thus, we term the technique cosine-distance virtual adversarial training (CD-VAT). In comparison to many existing SSL techniques, the unlabelled data does not have to come from the same set of classes (here speakers) as the labelled data. The effectiveness of CD-VAT is shown on the 2750+ hour VoxCeleb data set, where on a speaker verification task it achieves a reduction in equal error rate (EER) of 11.1% relative to a purely supervised baseline. This is 32.5% of the improvement that would be achieved from supervised training if the speaker labels for the unlabelled data were available.

The goal of this paper is text-independent speaker verification where utterances come from 'in the wild' videos and may contain irrelevant signal. While speaker verification is naturally a pair-wise problem, existing methods to produce the speaker embeddings are instance-wise. In this paper, we propose Cross Attentive Pooling (CAP) that utilizes the context information across the reference-query pair to generate utterance-level embeddings that contain the most discriminative information for the pair-wise matching problem. Experiments are performed on the VoxCeleb dataset in which our method outperforms comparable pooling strategies.