In this paper, we propose a Convolutional Neural Network (CNN) based speaker recognition model for extracting robust speaker embeddings. The embedding can be extracted efficiently with linear activation in the embedding layer. To understand how the speaker recognition model operates with text-independent input, we modify the structure to extract frame-level speaker embeddings from each hidden layer. We feed utterances from the TIMIT dataset to the trained network and use several proxy tasks to study the networks ability to represent speech input and differentiate voice identity. We found that the networks are better at discriminating broad phonetic classes than individual phonemes. In particular, frame-level embeddings that belong to the same phonetic classes are similar (based on cosine distance) for the same speaker. The frame level representation also allows us to analyze the networks at the frame level, and has the potential for other analyses to improve speaker recognition.
Deep generative models have achieved great success in unsupervised learning with the ability to capture complex nonlinear relationships between latent generating factors and observations. Among them, a factorized hierarchical variational autoencoder (FHVAE) is a variational inference-based model that formulates a hierarchical generative process for sequential data. Specifically, an FHVAE model can learn disentangled and interpretable representations, which have been proven useful for numerous speech applications, such as speaker verification, robust speech recognition, and voice conversion. However, as we will elaborate in this paper, the training algorithm proposed in the original paper is not scalable to datasets of thousands of hours, which makes this model less applicable on a larger scale. After identifying limitations in terms of runtime, memory, and hyperparameter optimization, we propose a hierarchical sampling training algorithm to address all three issues. Our proposed method is evaluated comprehensively on a wide variety of datasets, ranging from 3 to 1,000 hours and involving different types of generating factors, such as recording conditions and noise types. In addition, we also present a new visualization method for qualitatively evaluating the performance with respect to the interpretability and disentanglement. Models trained with our proposed algorithm demonstrate the desired characteristics on all the datasets.
The current trend in automatic speech recognition is to leverage large amounts of labeled data to train supervised neural network models. Unfortunately, obtaining data for a wide range of domains to train robust models can be costly. However, it is relatively inexpensive to collect large amounts of unlabeled data from domains that we want the models to generalize to. In this paper, we propose a novel unsupervised adaptation method that learns to synthesize labeled data for the target domain from unlabeled in-domain data and labeled out-of-domain data. We first learn without supervision an interpretable latent representation of speech that encodes linguistic and nuisance factors (e.g., speaker and channel) using different latent variables. To transform a labeled out-of-domain utterance without altering its transcript, we transform the latent nuisance variables while maintaining the linguistic variables. To demonstrate our approach, we focus on a channel mismatch setting, where the domain of interest is distant conversational speech, and labels are only available for close-talking speech. Our proposed method is evaluated on the AMI dataset, outperforming all baselines and bridging the gap between unadapted and in-domain models by over 77% without using any parallel data.
Speech recognizers trained on close-talking speech do not generalize to distant speech and the word error rate degradation can be as large as 40% absolute. Most studies focus on tackling distant speech recognition as a separate problem, leaving little effort to adapting close-talking speech recognizers to distant speech. In this work, we review several approaches from a domain adaptation perspective. These approaches, including speech enhancement, multi-condition training, data augmentation, and autoencoders, all involve a transformation of the data between domains. We conduct experiments on the AMI data set, where these approaches can be realized under the same controlled setting. These approaches lead to different amounts of improvement under their respective assumptions. The purpose of this paper is to quantify and characterize the performance gap between the two domains, setting up the basis for studying adaptation of speech recognizers from close-talking speech to distant speech. Our results also have implications for improving distant speech recognition.
In this paper, we propose a novel deep neural network architecture, Speech2Vec, for learning fixed-length vector representations of audio segments excised from a speech corpus, where the vectors contain semantic information pertaining to the underlying spoken words, and are close to other vectors in the embedding space if their corresponding underlying spoken words are semantically similar. The proposed model can be viewed as a speech version of Word2Vec. Its design is based on a RNN Encoder-Decoder framework, and borrows the methodology of skipgrams or continuous bag-of-words for training. Learning word embeddings directly from speech enables Speech2Vec to make use of the semantic information carried by speech that does not exist in plain text. The learned word embeddings are evaluated and analyzed on 13 widely used word similarity benchmarks, and outperform word embeddings learned by Word2Vec from the transcriptions.
Multimodal sensory data resembles the form of information perceived by humans for learning, and are easy to obtain in large quantities. Compared to unimodal data, synchronization of concepts between modalities in such data provides supervision for disentangling the underlying explanatory factors of each modality. Previous work leveraging multimodal data has mainly focused on retaining only the modality-invariant factors while discarding the rest. In this paper, we present a partitioned variational autoencoder (PVAE) and several training objectives to learn disentangled representations, which encode not only the shared factors, but also modality-dependent ones, into separate latent variables. Specifically, PVAE integrates a variational inference framework and a multimodal generative model that partitions the explanatory factors and conditions only on the relevant subset of them for generation. We evaluate our model on two parallel speech/image datasets, and demonstrate its ability to learn disentangled representations by qualitatively exploring within-modality and cross-modality conditional generation with semantics and styles specified by examples. For quantitative analysis, we evaluate the classification accuracy of automatically discovered semantic units. Our PVAE can achieve over 99% accuracy on both modalities.
We propose a process for investigating the extent to which sentence representations arising from neural machine translation (NMT) systems encode distinct semantic phenomena. We use these representations as features to train a natural language inference (NLI) classifier based on datasets recast from existing semantic annotations. In applying this process to a representative NMT system, we find its encoder appears most suited to supporting inferences at the syntax-semantics interface, as compared to anaphora resolution requiring world-knowledge. We conclude with a discussion on the merits and potential deficiencies of the existing process, and how it may be improved and extended as a broader framework for evaluating semantic coverage.
The back-propagation (BP) algorithm has been considered the de-facto method for training deep neural networks. It back-propagates errors from the output layer to the hidden layers in an exact manner using the transpose of the feedforward weights. However, it has been argued that this is not biologically plausible because back-propagating error signals with the exact incoming weights are not considered possible in biological neural systems. In this work, we propose a biologically plausible paradigm of neural architecture based on related literature in neuroscience and asymmetric BP-like methods. Specifically, we propose two bidirectional learning algorithms with trainable feedforward and feedback weights. The feedforward weights are used to relay activations from the inputs to target outputs. The feedback weights pass the error signals from the output layer to the hidden layers. Different from other asymmetric BP-like methods, the feedback weights are also plastic in our framework and are trained to approximate the forward activations. Preliminary results show that our models outperform other asymmetric BP-like methods on the MNIST and the CIFAR-10 datasets.
Although transfer learning has been shown to be successful for tasks like object and speech recognition, its applicability to question answering (QA) has yet to be well-studied. In this paper, we conduct extensive experiments to investigate the transferability of knowledge learned from a source QA dataset to a target dataset using two QA models. The performance of both models on a TOEFL listening comprehension test (Tseng et al., 2016) and MCTest (Richardson et al., 2013) is significantly improved via a simple transfer learning technique from MovieQA (Tapaswi et al., 2016). In particular, one of the models achieves the state-of-the-art on all target datasets; for the TOEFL listening comprehension test, it outperforms the previous best model by 7%. Finally, we show that transfer learning is helpful even in unsupervised scenarios when correct answers for target QA dataset examples are not available.
A reasonable approach for fact checking a claim involves retrieving potentially relevant documents from different sources (e.g., news websites, social media, etc.), determining the stance of each document with respect to the claim, and finally making a prediction about the claim's factuality by aggregating the strength of the stances, while taking the reliability of the source into account. Moreover, a fact checking system should be able to explain its decision by providing relevant extracts (rationales) from the documents. Yet, this setup is not directly supported by existing datasets, which treat fact checking, document retrieval, source credibility, stance detection and rationale extraction as independent tasks. In this paper, we support the interdependencies between these tasks as annotations in the same corpus. We implement this setup on an Arabic fact checking corpus, the first of its kind.