Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 10 of 10 for “"Speech processing systems"”.
-
The benefits of acoustic perceptual information for speech processing systems
… frame-synchronized framework has dominated many speech processing systems, such as ASR and AED targeting human speech activities. These systems have little consideration for the science behind speech and treat the task as a simple statistical classification. The framework also assumes each …
-
VLSI implementation of an efficient method for the computation of line spectral frequencies
<p>In speech coding applications using linear predictive techniques, the computation of line spectral frequencies (LSFs) from the predictor coefficients is an extremely computationally intensive task. The unique properties of the symmetric and antisymmetric polynomial roots limit the region which …
-
An analog VLSI vocal tract
… being used to improve performance in engineering systems. For example, silicon-cochlea-like models have led to improved speech recognition in noise and low-power cochlear-implant processors for the deaf. A promising approach to improve the naturalness of synthetic speech is to exploit bioinspired …
-
The pragmatics of the modern Greek grammatical system
… language interfaces, human-machine interaction, speech processing systems, and on-line language learning systems.
-
Use of linear predictive features and pattern recognition techniques to develop a vector quantization based blind SNR estimation system
… SNR of a communicated signal. Characteristics of speech signals can be used to develop non-intrusive methods for estimating the SNR of a signal. These methods do not require knowledge of the original speech signal for analysis.</p> <p>In this thesis a Vector Quantization (VQ) based pattern …
-
Adversarial Attacks on Natural Language and Speech Processing Models
… in computer vision, natural language processing, and speech processing. Despite their high performance, deep learning models are vulnerable to adversarial attacks. A deliberate and specific perturbation of a clean input sample can create an adversarial example, which, when processed by …
-
Machine Learning Solutions for Emotional Speech: Exploiting the Information of Individual Annotations
Speech emotion recognition solutions designed under controlled conditions do not translate well into real-life applications. The first step to address this issue is developing a large, natural emotional database. Current approaches used to collect spontaneous databases tend to provide unbalanced …
-
Incorporating non-verbal modalities in spoken language understanding for multimodal conversational systems
Thesis (Ph. D.)--Michigan State University. Computer Science, 2009