Dealing with limited and noisy data in ASR: a hybrid knowledge-based and statistical approach
نویسنده
چکیده
In this talk, I will focus on the importance of integrating knowledge of human speech production and speech perception mechanisms, and language-specific information with statisticallybased, data-driven approaches to develop robust and scalable automatic speech recognition (ASR) systems. As we will demonstrate, the need for such hybrid systems is especially critical when the ASR system is dealing with noisy data, when adaptation data are limited (for the case of speaker normalization and adaptation), and when dealing with accents.
منابع مشابه
An Improved Hybrid Model with Automated Lag Selection to Forecast Stock Market
Objective: In general, financial time series such as stock indexes have nonlinear, mutable and noisy behavior. Structural and statistical models and machine learning-based models are often unable to accurately predict series with such a behavior. Accordingly, the aim of the present study is to present a new hybrid model using the advantages of the GMDH method and Non-dominated Sorting Genetic A...
متن کاملDamage identification of structures using second-order approximation of Neumann series expansion
In this paper, a novel approach proposed for structural damage detection from limited number of sensors using extreme learning machine (ELM). As the number of sensors used to measure modal data is normally limited and usually are less than the number of DOFs in the finite element model, the model reduction approach should be used to match with incomplete measured mode shapes. The second-order a...
متن کاملA Hybrid Approach for Fuzzy Just-In-Time Flow Shop Scheduling with Limited Buffers and Deteriorating Jobs
This paper investigates the problem of just-in-time permutation flow shop scheduling with limited buffers and linear job deterioration in an uncertain environment. The fuzzy set theory is applied to describe this situation. A novel mixed-integer nonlinear program is presented to minimize the weighted sum of fuzzy earliness and tardiness penalties. Due to the computational complexities, the prop...
متن کاملAn ontological hybrid recommender system for dealing with cold start problem
Recommender Systems ( ) are expected to suggest the accurate goods to the consumers. Cold start is the most important challenge for RSs. Recent hybrid s combine and . We introduce an ontological hybrid RS where the ontology has been employed in its part while improving the ontology structure by its part. In this paper, a new hybrid approach is proposed based on the combination of demog...
متن کاملNoise adaptation for robust AURORA 2 noisy digit recognition using statistical data mapping
The mismatch between system training and operating conditions often has negative influences on automatic speech recognition (ASR) systems. Noise in the operating environments is commonly encountered. ASR model adaptation is an important way to enhance the system performance in noisy environments. This paper proposes a feature-based statistical data mapping (SDM) approach for robust noisy digit ...
متن کامل