Multi-domain adversarial training of neural network acoustic models for distant speech recognition

Mirsamadi, Seyedmahdad; Hansen, John H. L.

doi:10.1016/j.specom.2018.10.010

Cited by 21 publications

(9 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 1. Theoretically obtained SQNR for quasi- (16) where F is the total number of frames in which the signal is divided, SQNR(l). It should be noted that in the experimental analysis, M=240 samples per frame have been used, whereas the recorded benchmark test signal consists of about 4500 frames.…”

Section: Numerical Results and Analysismentioning

confidence: 99%

“…Quantization is the process of preparing a signal in digital domain and making it suitable for processing by a computer or any digital circuit [9]. Considering the growing interest in man-machine communication, speech and voice recognition is considered as important [37], [23][24], [36], [7], [16], [25]. Recent research and applications, which exploit neural networks, commonly incorporate quantization of weight coefficients and activation functions.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Simple Speech Transform Coding Scheme using Forward Adaptive Quantization for Discrete Input Signal

Perić

Tančić

Simic

et al. 2019

ITC

View full text Add to dashboard Cite

We propose a speech coding scheme based on the simple transform coding and forward adaptive quantization for discrete input signal processing in this paper. The quasi-logarithmic quantizer is applied to discretization of continuous input signal, i.e. for preparing discrete input. The application of forward adaptation based on the input signal variance provides more efficient bandwidth usage, whereas utilization of transform coding provides sub-sequences with more predictable signal characteristics that ensure higher quality of signal reconstruction at the receiving end. In order to provide additional compression, transform coding precedes adaptive quantization. As an objective measure of system performance we use signal-to-quantization-noise ratio. Sysem performance is discussed for two typical cases. In the first case, we consider that the information about continuous signal variance is available whereas the second case considers system performance estimation when we know only the information about discretized signal variance which means that there is a loss of input signal information. The main goal of such performance estimation comparison of the proposed speech signal coding model is to explore what is the objectivity of performance if we do not have information about a continuous source, which is a common phenomenon in digital systems.

show abstract

Section: Numerical Results and Analysismentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Simple Speech Transform Coding Scheme using Forward Adaptive Quantization for Discrete Input Signal

Perić

Tančić

Simic

et al. 2019

ITC

View full text Add to dashboard Cite

show abstract

“…Notations. In order to improve the recognition performance of the model, we consider using transfer learning [12][13][14][15]22]. Transfer learning is an ability of a system to recognize and apply knowledge and skills learned in previous domains/tasks to novel domains/tasks.…”

Section: Methodsmentioning

confidence: 99%

Basic Research on Ancient Bai Character Recognition Based on Mobile APP

Zhang

Lee

Gao

et al. 2021

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

Bai nationality has a long history and has its own language. Limited by the fact that there are fewer and fewer people who know the Bai language, the literature and culture of the Bai nationality begin to lose rapidly. In order to make the people who do not understand Bai characters can also read the ancient books of Bai nationality, this paper is based on the research of high-precision single character recognition model of Bai characters. First, with the help of Bai culture lovers and related scholars, we have constructed a data set of Bai characters, but limited by the need of expert knowledge, so the data set is limited in size. As a result, deep learning models with the nature of data hunger cannot get an ideal accuracy. In order to solve this issue, we propose to use the Chinese data set which also belongs to Sino-Tibetan language family to improve the recognition accuracy of Bai characters through transfer learning. In addition, we propose four transfer learning approaches: Direct Knowledge Transfer (DKT), Indirect Knowledge Transfer (IKT), Self-coding Knowledge Transfer (SCKT), and Self-supervised Knowledge Transfer (SSKT). Experiments show that our approaches greatly improve the recognition accuracy of Bai characters.

show abstract

“…Recently, adversarial training has achieved great success in speech recognition [27], [28], computer vision [29], [30] and natural language processing [31], [32]. Some works have also explored learning models based on confrontational training [33]- [35].…”

Section: B Gan-adversarial Trainingmentioning

confidence: 99%

Adversarial Training-Based Mean Bayesian Personalized Ranking for Recommender System

Wang

Han

2020

IEEE Access

View full text Add to dashboard Cite

Users' feedback information as the ground-truth has attracted a lot of attention in recommender systems. However, the feedback that could be contaminated by users' misoperations or malicious operations is probably not true in real scenarios. This work aims to develop a technique based on an improved Bayesian personalized ranking (BPR), called adversarial training-based mean Bayesian personalized ranking (AT-MBPR). In this method, we divide the feedback information into three categories based on the mean Bayesian personalized ranking (MBPR), then gain the implicit feedback from the mean and non-observed items of each user, following which, adversarial perturbations are added on the embedding vectors of the users and items by playing a minimax game to reduce the noise. The experiments demonstrate in five datasets that our approach outperforms the traditional BPR methods and state-of-the-art methods used for the recommendation. Our implementation is available at: https://github.com/ HanXia001/ Adversarial -Training-based-Mean-BPR-for-Recommender.

show abstract

Multi-domain adversarial training of neural network acoustic models for distant speech recognition

Cited by 21 publications

References 15 publications

Simple Speech Transform Coding Scheme using Forward Adaptive Quantization for Discrete Input Signal

Simple Speech Transform Coding Scheme using Forward Adaptive Quantization for Discrete Input Signal

Basic Research on Ancient Bai Character Recognition Based on Mobile APP

Adversarial Training-Based Mean Bayesian Personalized Ranking for Recommender System

Contact Info

Product

Resources

About