In moving jammer environments, the performance of conventional adaptive beamformer is severely degraded and the robust adaptive beamformer requires additional sensors to obtain desired performances. Therefore, it is necessary to develop efficient algorithm without any additional requirement of the number of sensors, etc. In this paper, we introduce a fast adaptive algorithm with variable forgetting factor, which does not have any additional requirements. From the computer simulations, we obtain the better performances than those of other techniques for the arrays with various aperture lengths.

A theoretical formula that is based on the geometrical theory of diffraction (GTD) is proposed for computing sound diffraction by multiple wedges, barriers, and polygonal-like shapes. The formula can treat both convex and concave edges, where edges mayor may not be inter-connected. Comparisons of theoretical predictions with other results done by the BEM or experiments for scaled model confirm the accuracy of the present formula. Numerical examples such as double wedges and doubly inclined barrier show that when there exist several diffraction paths for given source and receiver positions, the insertion loss is dominated by the diffraction associated with the shortest propagation path.

In this paper, we propose a style-specific language model adaptation scheme using n-gram based tf*idf similarity for Korean spontaneous speech recognition. Korean spontaneous speech shows especially different style-specific characteristics such as filled pauses, word omission, and contraction, which are related to function words and depend on preceding or following words. To reflect these style-specific characteristics and overcome insufficient data for training language model, we estimate in-domain dependent n-gram model by relevance weighting of out-of-domain text data according to their n-. gram based tf*idf similarity, in which in-domain language model include disfluency model. Recognition results show that n-gram based tf*idf similarity weighting effectively reflects style difference.

Despite abundant research outcomes of blind source separation (BSS) in many types of simulated environments, their performances are still not satisfactory to be applied to the real environments. The major obstacle may seem the finite filter length of the assumed mixing model and the nonlinear sensor noises. This paper presents a two-step speech enhancement method with multiple microphone inputs. The first step performs a frequency-domain BSS algorithm to produce multiple outputs without any prior knowledge of the mixed source signals. The second step further removes the remaining cross-channel interference by a spectral cancellation approach using a probabilistic source absence/presence detection technique. The desired primary source is detected every frame of the signal, and the secondary source is estimated in the power spectral domain using the other BSS output as a reference interfering source. Then the estimated secondary source is subtracted to reduce the cross-channel interference. Our experimental results show good separation enhancement performances on the real recordings of speech and music signals compared to the conventional BSS methods.

In the East Sea of Korea, the ocean environments are known to have strong variations in space and time. Their effects are very important factors in sound propagation and sonar performance. We consider the environmental factors such as eddies and thermal fronts affecting underwater sound propagation and target detection performance by sonars. Unfortunately, however, the detailed structure of eddies is usually difficult to understand by using the sea surface temperatures from infrared images alone or a few profiles from the CTD (conductivity, temperature and depth) castings. The temperature fields of eddy and thermal front are simulated with typical patterns of those obtained from several observations. This paper delivers the overviews of environments and acoustic models with their simulation results on sonar performance.