Thesis (M.S.)--Worcester Polytechnic Institute. / Keywords: multimedia; content adaptation; scalable web server. Includes bibliographical references (p. 68-74).

published_or_final_version / abstract / Computer Science / Master / Master of Philosophy

The past decade has seen extensive research on audio classification algorithms which playa key role in multimedia applications, such as the retrieval of audio information from an audio or audiovisual database. However, the effect of background noise on the performance of classification has not been widely investigated. Motivated by the noise-suppression property of the early auditory (EA) model presented by Wang and Shamma, we seek in this thesis to further investigate this property and to develop improved algorithms for audio classification in the presence of background noise. / With respect to the limitation of the original analysis, a better yet mathematically tractable approximation approach is first proposed wherein the Gaussian cumulative distribution function is used to derive a new closed-form expression of the auditory spectrum at the output of the EA model, and to conduct relevant analysis. Considering the computational complexity of the original EA model, a simplified auditory spectrum is proposed, wherein the underlying analysis naturally leads to frequency-domain approximation for further reduction in the computational complexity. Based on this time-domain approximation, a simplified FFT-based spectrum is proposed wherein a local spectral self-normalization is implemented. An improved implementation of this spectrum is further proposed to calculate a so-called FFT-based auditory spectrum, which allows more flexibility in the extraction of noise-robust audio features. / To evaluate the performance of the above FFT-based spectra, speech/music/noise and noise/non-noise classification experiments are conducted wherein a support vector machine algorithm (SVMstruct) and a decision tree learning algorithm (C4.5) are used as the classifiers. Several features are used for the classification, including the conventional mel-frequency cepstral coefficient (MFCC) features as well as DCT-based and spectral features derived from the proposed FFT-based spectra. Compared to the conventional features, the auditory-related features show more robust performance in mismatched test cases. Test results also indicate that the performance of the proposed FFT-based auditory spectrum is slightly better than that of the original auditory spectrum, while its computational complexity is reduced by an order of magnitude. / Finally, to further explore the proposed FFT-based auditory spectrum from a practical audio classification perspective, a floating-point DSP implementation is developed and optimized on the TMS320C6713 DSP Starter Kit (DSK) from Texas Instruments.

The comprehension of illustrations and text was studied from a cognitive discourse processing perspective. Typically, learners must construct conceptual knowledge representations that integrate different types of information from diverse sources and modalities (e.g., text and illustrations). Currently, little is known about how such integrative processing works in "multimedia" learning situations. This study focused on the semantic representations that low prior knowledge learners constructed as they read from text and static illustrations presenting multiple types of information (structural, functional, and energy) describing a functional system. Both the text and the illustrations were modified so that structural information would be highlighted over other information types. / Participants were twenty-four undergraduate engineering students who had little prior knowledge of the target domain (the human visual system), but were experienced in learning about functional systems using texts and illustrations. Six students were randomly assigned to each of four presentation conditions: (a) text only, (b) illustrations only, (c) text with controlled access to illustrations, or (d) text with free access to illustrations. / The materials were presented individually in a computer environment which recorded and timed all information accessed. Participants provided on-line interpretations as they read, post-input verbal and visual free recalls of the materials, and responses to integrative comprehension questions. Planned comparisons were used to contrast: (a) the two text with illustrations groups, (b) the combined text with illustrations groups to the text only group, (c) the text with illustrations groups to the illustrations only group, and (d) the processing of information which was privileged (structure) to other information. / The results indicated that the text and illustrations each provided mutually constraining information that functioned together to support comprehension. Illustrations aided the construction and elaboration of mental models by providing an external context that supported more active conceptual processing and integration of information. Text aided both literal and high level comprehension by communicating the meaning of illustrations and signaling what information was important. These results support perspectives on situated learning which emphasize the role that discourse plays in comprehending knowledge in environments involving more than one external information source. Future research is recommended to extend such findings to other populations of learners and materials.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.