arXiv subjects
Publications and source records attributed to Zheng-Hua Tan.
In this paper we derive a PAC-Bayesian error bound for autonomous stochastic LTI state-space models. The motivation for deriving such error bounds is that they will allow deriving similar error bounds for more general dynamical systems, including recurrent neural networks. In turn, PACBayesian error bounds are known to be useful for analyzing machine learning algorithms and for deriving new ones.
In this letter, we propose a vocal tract length (VTL) perturbation method for text-dependent speaker verification (TD-SV), in which a set of TD-SV systems are trained, one for each VTL factor, and score-level fusion is applied to make a final decision. Next, we explore the bottleneck (BN) feature extracted by training deep neural networks with a self-supervised objective, autoregressive predictive coding (APC), for TD-SV and compare it with the well-studied speaker-discriminant BN feature. The proposed VTL method is then applied to APC and speaker-discriminant BN features. In the end, we combine the VTL perturbation systems trained on MFCC and the two BN features in the score domain. Experiments are performed on the RedDots challenge 2016 database of TD-SV using short utterances with Gaussian mixture model-universal background model and i-vector techniques. Results show the proposed methods significantly outperform the baselines.
Speech enhancement and speech separation are two related tasks, whose purpose is to extract either one or more target speech signals, respectively, from a mixture of sounds generated by several sources. Traditionally, these tasks have been tackled using signal processing and machine learning techniques applied to the available acoustic signals. Since the visual aspect of speech is essentially unaffected by the acoustic environment, visual information from the target speakers, such as lip movements and facial expressions, has also been used for speech enhancement and speech separation systems. In order to efficiently fuse acoustic and visual information, researchers have exploited the flexibility of data-driven approaches, specifically deep learning, achieving strong performance. The ceaseless proposal of a large number of techniques to extract features and fuse multimodal information has highlighted the need for an overview that comprehensively describes and discusses audio-visual speech enhancement and separation based on deep learning. In this paper, we provide a systematic survey of this research topic, focusing on the main elements that characterise the systems in the literature: acoustic features; visual features; deep learning methods; fusion techniques; training targets and objective functions. In addition, we review deep-learning-based methods for speech reconstruction from silent videos and audio-visual sound source separation for non-speech signals, since these methods can be more or less directly applied to audio-visual speech enhancement and separation. Finally, we survey commonly employed audio-visual speech datasets, given their central role in the development of data-driven approaches, and evaluation methods, because they are generally used to compare different systems and determine their performance.
In this paper, we present a deep-learning-based framework for audio-visual speech inpainting, i.e., the task of restoring the missing parts of an acoustic speech signal from reliable audio context and uncorrupted visual information. Recent work focuses solely on audio-only methods and generally aims at inpainting music signals, which show highly different structure than speech. Instead, we inpaint speech signals with gaps ranging from 100 ms to 1600 ms to investigate the contribution that vision can provide for gaps of different duration. We also experiment with a multi-task learning approach where a phone recognition task is learned together with speech inpainting. Results show that the performance of audio-only speech inpainting approaches degrades rapidly when gaps get large, while the proposed audio-visual approach is able to plausibly restore missing information. In addition, we show that multi-task learning is effective, although the largest contribution to performance comes from vision.
In this paper, we propose a novel method that trains pass-phrase specific deep neural network (PP-DNN) based auto-encoders for creating augmented data for text-dependent speaker verification (TD-SV). Each PP-DNN auto-encoder is trained using the utterances of a particular pass-phrase available in the target enrollment set with two methods: (i) transfer learning and (ii) training from scratch. Next, feature vectors of a given utterance are fed to the PP-DNNs and the output from each PP-DNN at frame-level is considered one new set of generated data. The generated data from each PP-DNN is then used for building a TD-SV system in contrast to the conventional method that considers only the evaluation data available. The proposed approach can be considered as the transformation of data to the pass-phrase specific space using a non-linear transformation learned by each PP-DNN. The method develops several TD-SV systems with the number equal to the number of PP-DNNs separately trained for each pass-phrases for the evaluation. Finally, the scores of the different TD-SV systems are fused for decision making. Experiments are conducted on the RedDots challenge 2016 database for TD-SV using short utterances. Results show that the proposed method improves the performance for both conventional cepstral feature and deep bottleneck feature using both Gaussian mixture model - universal background model (GMM-UBM) and i-vector framework.
One of the beyond-5G developments that is often highlighted is the integration of wireless communication and radio sensing. This paper addresses the potential of communication-sensing integration of Large Intelligent Surfaces (LIS) in an exemplary Industry 4.0 scenario. Besides the potential for high throughput and efficient multiplexing of wireless links, an LIS can offer a high-resolution rendering of the propagation environment. This is because, in an indoor setting, it can be placed in proximity to the sensed phenomena, while the high resolution is offered by densely spaced tiny antennas deployed over a large area. By treating an LIS as a radio image of the environment, we develop sensing techniques that leverage the usage of computer vision combined with machine learning. We test these methods for a scenario where we need to detect whether an industrial robot deviates from a predefined route. The results show that the LIS-based sensing offers high precision and has a high application potential in indoor industrial environments.
The loss function is a key component in deep learning models. A commonly used loss function for classification is the cross entropy loss, which is a simple yet effective application of information theory for classification problems. Based on this loss, many other loss functions have been proposed,~\emph{e.g.}, by adding intra-class and inter-class constraints to enhance the discriminative ability of the learned features. However, these loss functions fail to consider the connections between the feature distribution and the model structure. Aiming at addressing this problem, we propose a channel correlation loss (CC-Loss) that is able to constrain the specific relations between classes and channels as well as maintain the intra-class and the inter-class separability. CC-Loss uses a channel attention module to generate channel attention of features for each sample in the training stage. Next, an Euclidean distance matrix is calculated to make the channel attention vectors associated with the same class become identical and to increase the difference between different classes. Finally, we obtain a feature embedding with good intra-class compactness and inter-class separability.Experimental results show that two different backbone models trained with the proposed CC-Loss outperform the state-of-the-art loss functions on three image classification datasets.
Applying x-vectors for speaker verification has recently attracted great interest, with the focus being on text-independent speaker verification. In this paper, we study x-vectors for text-dependent speaker verification (TD-SV), which remains unexplored. We further investigate the impact of the different bottleneck (BN) features on the performance of x-vectors, including the recently-introduced time-contrastive-learning (TCL) BN features and phone-discriminant BN features. TCL is a weakly supervised learning approach that constructs training data by uniformly partitioning each utterance into a predefined number of segments and then assigning each segment a class label depending on their position in the utterance. We also compare TD-SV performance for different modeling techniques, including the Gaussian mixture models-universal background model (GMM-UBM), i-vector, and x-vector. Experiments are conducted on the RedDots 2016 challenge database. It is found that the type of features has a marginal impact on the performance of x-vectors with the TCL BN feature achieving the lowest equal error rate, while the impact of features is significant for i-vector and GMM-UBM. The fusion of x-vector and i-vector systems gives a large gain in performance. The GMM-UBM technique shows its advantage for TD-SV using short utterances.
Both acoustic and visual information influence human perception of speech. For this reason, the lack of audio in a video sequence determines an extremely low speech intelligibility for untrained lip readers. In this paper, we present a way to synthesise speech from the silent video of a talker using deep learning. The system learns a mapping function from raw video frames to acoustic features and reconstructs the speech with a vocoder synthesis algorithm. To improve speech reconstruction performance, our model is also trained to predict text information in a multi-task learning fashion and it is able to simultaneously reconstruct and recognise speech in real time. The results in terms of estimated speech quality and intelligibility show the effectiveness of our method, which exhibits an improvement over existing video-to-speech approaches.
In this work, we present the system description of the UIAI entry for the short-duration speaker verification (SdSV) challenge 2020. Our focus is on Task 1 dedicated to text-dependent speaker verification. We investigate different feature extraction and modeling approaches for automatic speaker verification (ASV) and utterance verification (UV). We have also studied different fusion strategies for combining UV and ASV modules. Our primary submission to the challenge is the fusion of seven subsystems which yields a normalized minimum detection cost function (minDCF) of 0.072 and an equal error rate (EER) of 2.14% on the evaluation set. The single system consisting of a pass-phrase identification based model with phone-discriminative bottleneck features gives a normalized minDCF of 0.118 and achieves 19% relative improvement over the state-of-the-art challenge baseline.
Data augmentation is commonly used for generating additional data from the available training data to achieve a robust estimation of the parameters of complex models like the one for speaker verification (SV), especially for under-resourced applications. SV involves training speaker-independent (SI) models and speaker-dependent models where speakers are represented by models derived from an SI model using the training data for the particular speaker during the enrollment phase. While data augmentation for training SI models is well studied, data augmentation for speaker enrollment is rarely explored. In this paper, we propose the use of data augmentation methods for generating extra data to empower speaker enrollment. Each data augmentation method generates a new data set. Two strategies of using the data sets are explored: the first one is to training separate systems and fuses them at the score level and the other is to conduct multi-conditional training. Furthermore, we study the effect of data augmentation under noisy conditions. Experiments are performed on RedDots challenge 2016 database, and the results validate the effectiveness of the proposed methods.
Despite their great performance over the years, handcrafted speech features are not necessarily optimal for any particular speech application. Consequently, with greater or lesser success, optimal filterbank learning has been studied for different speech processing tasks. In this paper, we fill in a gap by exploring filterbank learning for keyword spotting (KWS). Two approaches are examined: filterbank matrix learning in the power spectral domain and parameter learning of a psychoacoustically-motivated gammachirp filterbank. Filterbank parameters are optimized jointly with a modern deep residual neural network-based KWS back-end. Our experimental results reveal that, in general, there are no statistically significant differences, in terms of KWS accuracy, between using a learned filterbank and handcrafted speech features. Thus, while we conclude that the latter are still a wise choice when using modern KWS back-ends, we also hypothesize that this could be a symptom of information redundancy, which opens up new research possibilities in the field of small-footprint KWS.
A deep neural network of multiple nonlinear layers forms a large function space, which can easily lead to overfitting when it encounters small-sample data. To mitigate overfitting in small-sample classification, learning more discriminative features from small-sample data is becoming a new trend. To this end, this paper aims to find a subspace of neural networks that can facilitate a large decision margin. Specifically, we propose the Orthogonal Softmax Layer (OSL), which makes the weight vectors in the classification layer remain orthogonal during both the training and test processes. The Rademacher complexity of a network using the OSL is only $\frac{1}{K}$, where $K$ is the number of classes, of that of a network using the fully connected classification layer, leading to a tighter generalization error bound. Experimental results demonstrate that the proposed OSL has better performance than the methods used for comparison on four small-sample benchmark datasets, as well as its applicability to large-sample datasets. Codes are available at: https://github.com/dongliangchang/OSLNet.
Many deep learning-based speech enhancement algorithms are designed to minimize the mean-square error (MSE) in some transform domain between a predicted and a target speech signal. However, optimizing for MSE does not necessarily guarantee high speech quality or intelligibility, which is the ultimate goal of many speech enhancement algorithms. Additionally, only little is known about the impact of the loss function on the emerging class of time-domain deep learning-based speech enhancement systems. We study how popular loss functions influence the performance of deep learning-based speech enhancement systems. First, we demonstrate that perceptually inspired loss functions might be advantageous if the receiver is the human auditory system. Furthermore, we show that the learning rate is a crucial design parameter even for adaptive gradient-based optimizers, which has been generally overlooked in the literature. Also, we found that waveform matching performance metrics must be used with caution as they in certain situations can fail completely. Finally, we show that a loss function based on scale-invariant signal-to-distortion ratio (SI-SDR) achieves good general performance across a range of popular speech enhancement evaluation metrics, which suggests that SI-SDR is a good candidate as a general-purpose loss function for speech enhancement systems.
This paper proposes a deep learning-based method for learning joint context-content embeddings (JCCE) with a view to context-aware recommendations, and demonstrate its application in the television domain. JCCE builds on recent progress within latent representations for recommendation and deep metric learning. The model effectively groups viewing situations and associated consumed content, based on supervision from 2.7 million viewing events. Experiments confirm the recommendation ability of JCCE, achieving improvements when compared to state-of-the-art methods. Furthermore, the approach shows meaningful structures in the learned representations that can be used to gain valuable insights of underlying factors in the relationship between contextual settings and content properties.
Machine Learning systems are vulnerable to adversarial attacks and will highly likely produce incorrect outputs under these attacks. There are white-box and black-box attacks regarding to adversary's access level to the victim learning algorithm. To defend the learning systems from these attacks, existing methods in the speech domain focus on modifying input signals and testing the behaviours of speech recognizers. We, however, formulate the defense as a classification problem and present a strategy for systematically generating adversarial example datasets: one for white-box attacks and one for black-box attacks, containing both adversarial and normal examples. The white-box attack is a gradient-based method on Baidu DeepSpeech with the Mozilla Common Voice database while the black-box attack is a gradient-free method on a deep model-based keyword spotting system with the Google Speech Command dataset. The generated datasets are used to train a proposed Convolutional Neural Network (CNN), together with cepstral features, to detect adversarial examples. Experimental results show that, it is possible to accurately distinct between adversarial and normal examples for known attacks, in both single-condition and multi-condition training settings, while the performance degrades dramatically for unknown attacks. The adversarial datasets and the source code are made publicly available.
Home entertainment systems feature in a variety of usage scenarios with one or more simultaneous users, for whom the complexity of choosing media to consume has increased rapidly over the last decade. Users' decision processes are complex and highly influenced by contextual settings, but data supporting the development and evaluation of context-aware recommender systems are scarce. In this paper we present a dataset of self-reported TV consumption enriched with contextual information of viewing situations. We show how choice of genre associates with, among others, the number of present users and users' attention levels. Furthermore, we evaluate the performance of predicting chosen genres given different configurations of contextual information, and compare the results to contextless predictions. The results suggest that including contextual features in the prediction cause notable improvements, and both temporal and social context show significant contributions.
Keyword spotting (KWS) is experiencing an upswing due to the pervasiveness of small electronic devices that allow interaction with them via speech. Often, KWS systems are speaker-independent, which means that any person --user or not-- might trigger them. For applications like KWS for hearing assistive devices this is unacceptable, as only the user must be allowed to handle them. In this paper we propose KWS for hearing assistive devices that is robust to external speakers. A state-of-the-art deep residual network for small-footprint KWS is regarded as a basis to build upon. By following a multi-task learning scheme, this system is extended to jointly perform KWS and users' own-voice/external speaker detection with a negligible increase in the number of parameters. For experiments, we generate from the Google Speech Commands Dataset a speech corpus emulating hearing aids as a capturing device. Our results show that this multi-task deep residual network is able to achieve a KWS accuracy relative improvement of around 32% with respect to a system that does not deal with external speakers.