Abstract: The current speech interfaces in many military
applications may be adequate for native speakers. However,
the recognition rate drops quite a lot for non-native speakers
(people with foreign accents). This is mainly because the nonnative
speakers have large temporal and intra-phoneme
variations when they pronounce the same words. This
problem is also complicated by the presence of large
environmental noise such as tank noise, helicopter noise, etc.
In this paper, we proposed a novel continuous acoustic feature
adaptation algorithm for on-line accent and environmental
adaptation. Implemented by incremental singular value
decomposition (SVD), the algorithm captures local acoustic
variation and runs in real-time. This feature-based adaptation
method is then integrated with conventional model-based
maximum likelihood linear regression (MLLR) algorithm.
Extensive experiments have been performed on the NATO
non-native speech corpus with baseline acoustic model trained
on native American English. The proposed feature-based
adaptation algorithm improved the average recognition
accuracy by 15%, while the MLLR model based adaptation
achieved 11% improvement. The corresponding word error
rate (WER) reduction was 25.8% and 2.73%, as compared to
that without adaptation. The combined adaptation achieved
overall recognition accuracy improvement of 29.5%, and
WER reduction of 31.8%, as compared to that without
adaptation.
Abstract: This paper presents the first reflexions about Margaret Mascarenhas-s novel, “Skin", based on post-colonial critic perception of History and its agents. By doing so, this study will put light on a literary corpus of Indian Literatures: the Goan Literature whose cultural basis creates an unique historiographic metafiction conducted by different characters that one by one plays the narrator role.
Abstract: Information Retrieval has the objective of studying
models and the realization of systems allowing a user to find the
relevant documents adapted to his need of information. The
information search is a problem which remains difficult because the
difficulty in the representing and to treat the natural languages such
as polysemia. Intentional Structures promise to be a new paradigm to
extend the existing documents structures and to enhance the different
phases of documents process such as creation, editing, search and
retrieval. The intention recognition of the author-s of texts can reduce
the largeness of this problem. In this article, we present intentions
recognition system is based on a semi-automatic method of
extraction the intentional information starting from a corpus of text.
This system is also able to update the ontology of intentions for the
enrichment of the knowledge base containing all possible intentions
of a domain. This approach uses the construction of a semi-formal
ontology which considered as the conceptualization of the intentional
information contained in a text. An experiments on scientific
publications in the field of computer science was considered to
validate this approach.
Abstract: Speech corpus is one of the major components in a
Speech Processing System where one of the primary requirements
is to recognize an input sample. The quality and details captured
in speech corpus directly affects the precision of recognition. The
current work proposes a platform for speech corpus generation using
an adaptive LMS filter and LPC cepstrum, as a part of an ANN
based Speech Recognition System which is exclusively designed to
recognize isolated numerals of Assamese language- a major language
in the North Eastern part of India. The work focuses on designing an
optimal feature extraction block and a few ANN based cooperative
architectures so that the performance of the Speech Recognition
System can be improved.