Modern speech synthesis techniques can produce natural-sounding speech given sufficient high-quality data and compute resources. However, such data is not readily available for many languages. This paper focuses on speech synthesis for low-resourced African languages, from corpus creation to sharing and deploying the Text-to-Speech (TTS) systems. We first create a set of general-purpose instructions on building speech synthesis systems with minimum technological resources and subject-matter expertise. Next, we create new datasets and curate datasets from "found" data (existing recordings) through a participatory approach while considering accessibility, quality, and breadth. We demonstrate that we can develop synthesizers that generate intelligible speech with 25 minutes of created speech, even when recorded in suboptimal environments. Finally, we release the speech data, code, and trained voices for 12 African languages to support researchers and developers.
Integrating vision and language has gained notable attention following the success of pretrained language models. Despite that, a fraction of emerging multimodal models is suitable for text generation conditioned on images. This minority is typically developed and evaluated for image captioning, a text generation task conditioned solely on images with the goal to describe what is explicitly visible in an image. In this paper, we take a step back and ask: How do these models work for more complex generative tasks, conditioned on both text and images? Are models based on joint multimodal pretraining, visually adapted pretrained language models, or models that combine these two approaches, more promising for such tasks? We address these questions in the context of self-rationalization (jointly generating task labels/answers and free-text explanations) of three tasks: (i) visual question answering in VQA-X, (ii) visual commonsense reasoning in VCR, and (iii) visual-textual entailment in E-SNLI-VE. We show that recent advances in each modality, CLIP image representations and scaling of language models, do not consistently improve multimodal self-rationalization of tasks with multimodal inputs. We also observe that no model type works universally the best across tasks/datasets and finetuning data sizes. Our findings call for a backbone modelling approach that can be built on to advance text generation from images and text beyond image captioning.
Most of the research on data-driven speech representation learning has focused on raw audios in an end-to-end manner, paying little attention to their internal phonological or gestural structure. This work, investigating the speech representations derived from articulatory kinematics signals, uses a neural implementation of convolutive sparse matrix factorization to decompose the articulatory data into interpretable gestures and gestural scores. By applying sparse constraints, the gestural scores leverage the discrete combinatorial properties of phonological gestures. Phoneme recognition experiments were additionally performed to show that gestural scores indeed code phonological information successfully. The proposed work thus makes a bridge between articulatory phonology and deep neural networks to leverage informative, intelligible, interpretable,and efficient speech representations.
As Automatic Speech Processing (ASR) systems are getting better, there is an increasing interest of using the ASR output to do downstream Natural Language Processing (NLP) tasks. However, there are few open source toolkits that can be used to generate reproducible results on different Spoken Language Understanding (SLU) benchmarks. Hence, there is a need to build an open source standard that can be used to have a faster start into SLU research. We present ESPnet-SLU, which is designed for quick development of spoken language understanding in a single framework. ESPnet-SLU is a project inside end-to-end speech processing toolkit, ESPnet, which is a widely used open-source standard for various speech processing tasks like ASR, Text to Speech (TTS) and Speech Translation (ST). We enhance the toolkit to provide implementations for various SLU benchmarks that enable researchers to seamlessly mix-and-match different ASR and NLU models. We also provide pretrained models with intensively tuned hyper-parameters that can match or even outperform the current state-of-the-art performances. The toolkit is publicly available at https://github.com/espnet/espnet.
Speech processing systems currently do not support the vast majority of languages, in part due to the lack of data in low-resource languages. Cross-lingual transfer offers a compelling way to help bridge this digital divide by incorporating high-resource data into low-resource systems. Current cross-lingual algorithms have shown success in text-based tasks and speech-related tasks over some low-resource languages. However, scaling up speech systems to support hundreds of low-resource languages remains unsolved. To help bridge this gap, we propose a language similarity approach that can efficiently identify acoustic cross-lingual transfer pairs across hundreds of languages. We demonstrate the effectiveness of our approach in language family classification, speech recognition, and speech synthesis tasks.
Code-switching (CS), a ubiquitous phenomenon due to the ease of communication it offers in multilingual communities still remains an understudied problem in language processing. The primary reasons behind this are: (1) minimal efforts in leveraging large pretrained multilingual models, and (2) the lack of annotated data. The distinguishing case of low performance of multilingual models in CS is the intra-sentence mixing of languages leading to switch points. We first benchmark two sequence labeling tasks -- POS and NER on 4 different language pairs with a suite of pretrained models to identify the problems and select the best performing model, char-BERT, among them (addressing (1)). We then propose a self training method to repurpose the existing pretrained models using a switch-point bias by leveraging unannotated data (addressing (2)). We finally demonstrate that our approach performs well on both tasks by reducing the gap between the switch point performance while retaining the overall performance on two distinct language pairs in both the tasks. Our code is available here: https://github.com/PC09/EMNLP2021-Switch-Point-biased-Self-Training.
Language models (LMs) for text data have been studied extensively for their usefulness in language generation and other downstream tasks. However, language modelling purely in the speech domain is still a relatively unexplored topic, with traditional speech LMs often depending on auxiliary text LMs for learning distributional aspects of the language. For the English language, these LMs treat words as atomic units, which presents inherent challenges to language modelling in the speech domain. In this paper, we propose a novel LSTM-based generative speech LM that is inspired by the CBOW model and built on linguistic units including syllables and phonemes. This offers better acoustic consistency across utterances in the dataset, as opposed to single melspectrogram frames, or whole words. With a limited dataset, orders of magnitude smaller than that required by contemporary generative models, our model closely approximates babbling speech. We show the effect of training with auxiliary text LMs, multitask learning objectives, and auxiliary articulatory features. Through our experiments, we also highlight some well known, but poorly documented challenges in training generative speech LMs, including the mismatch between the supervised learning objective with which these models are trained such as Mean Squared Error (MSE), and the true objective, which is speech quality. Our experiments provide an early indication that while validation loss and Mel Cepstral Distortion (MCD) are not strongly correlated with generated speech quality, traditional text language modelling metrics like perplexity and next-token-prediction accuracy might be.
Building Spoken Language Understanding (SLU) systems that do not rely on language specific Automatic Speech Recognition (ASR) is an important yet less explored problem in language processing. In this paper, we present a comparative study aimed at employing a pre-trained acoustic model to perform SLU in low resource scenarios. Specifically, we use three different embeddings extracted using Allosaurus, a pre-trained universal phone decoder: (1) Phone (2) Panphone, and (3) Allo embeddings. These embeddings are then used in identifying the spoken intent. We perform experiments across three different languages: English, Sinhala, and Tamil each with different data sizes to simulate high, medium, and low resource scenarios. Our system improves on the state-of-the-art (SOTA) intent classification accuracy by approximately 2.11% for Sinhala and 7.00% for Tamil and achieves competitive results on English. Furthermore, we present a quantitative analysis of how the performance scales with the number of training examples used per intent.
Speech summarization is typically performed by using a cascade of speech recognition and text summarization models. End-to-end modeling of speech summarization models is challenging due to memory and compute constraints arising from long input audio sequences. Recent work in document summarization has inspired methods to reduce the complexity of self-attentions, which enables transformer models to handle long sequences. In this work, we introduce a single model optimized end-to-end for speech summarization. We apply the restricted self-attention technique from text-based models to speech models to address the memory and compute constraints. We demonstrate that the proposed model learns to directly summarize speech for the How-2 corpus of instructional videos. The proposed end-to-end model outperforms the previously proposed cascaded model by 3 points absolute on ROUGE. Further, we consider the spoken language understanding task of predicting concepts from speech inputs and show that the proposed end-to-end model outperforms the cascade model by 4 points absolute F-1.
Decomposable tasks are complex and comprise of a hierarchy of sub-tasks. Spoken intent prediction, for example, combines automatic speech recognition and natural language understanding. Existing benchmarks, however, typically hold out examples for only the surface-level sub-task. As a result, models with similar performance on these benchmarks may have unobserved performance differences on the other sub-tasks. To allow insightful comparisons between competitive end-to-end architectures, we propose a framework to construct robust test sets using coordinate ascent over sub-task specific utility functions. Given a dataset for a decomposable task, our method optimally creates a test set for each sub-task to individually assess sub-components of the end-to-end model. Using spoken language understanding as a case study, we generate new splits for the Fluent Speech Commands and Snips SmartLights datasets. Each split has two test sets: one with held-out utterances assessing natural language understanding abilities, and one with held-out speakers to test speech processing skills. Our splits identify performance gaps up to 10% between end-to-end systems that were within 1% of each other on the original test sets. These performance gaps allow more realistic and actionable comparisons between different architectures, driving future model development. We release our splits and tools for the community.