A spectral atmospheric general circulation model called AFES (AGCM for Earth Simulator) was developed and optimized for the architecture of the Earth Simulator (ES). The ES is a massively parallel vector supercomputer that consists of 640 processor nodes interconnected by a single stage crossbar network with its total peak performance of 40.96 Tflops was achieved for a high resolution simulation (T1279L96) with AFES by utilizing the full 640-node configuration of the ES. The resulting computing efficiency is 64.9% of the peak performance, well surpassing that of conventional weather/climate applications having just 25-50% efficiency even on vector parallel computers. This remarkable performance proves the effectiveness of the ES as a viable means for practical applications.

This paper proposes instantaneous environment adaptation techniques for both additive noise and channel distortion based on the fast PMC (FPMC) and the MAP-CMS methods. The instantaneous adaptation techniques enable a recognizer to improve recognition on a single sentence that is used for the adaptation in real-time. The key innovations enabling the system to achieve the instantaneous adaptation are: (1) a cepstral mean subtraction method based on maximum a posteriori estimation (MAP-CMS), (2) real-time implementation of the fast parallel model combination (PMC) that we proposed previously, (3) utilization of multi-pass search, and (4) a new combination method of MAP-CMS and FPMC to solve the problem of both channel distortion and additive noise. Experiment results showed that the proposed methods enabled the system to perform recognition and adaptation simultaneously nearly in real-time and obtained good improvements in performance.

We describe a fast speech recognition algorithm under a noisy environment. To achieve accurate and fast speech recognition under a noisy environment, a very fast speech recognition algorithm with well-adapted model against the noisy environment is required. First, for the model adaptation, we propose the MCMS-PMC: an integration of parallel model combination (PMC) and modified cepstral mean subtraction (MCMS) which estimates the cepstrum mean by taking account of the additive noise. Then, for the fast speech recognition, we propose new techniques to create the noise-adapted scalar quantized codebook in order to introduce the MCMS-PMC into the IDMM+SQ, which we proposed previously as a fast speech recognition algorithm using the scalar quantization approach. Finally, an effect of the proposed method is shown through the speaker-independent telephone-bandwidth continuous speech recognition experiment.

In this paper, we propose a high speed output probability computation algorithm for multi-mixture continuous HMM. In the algorithm, we adopted the following three techniques: 1) independent dimension multi-mixture computation (IDMM), 2) scalar quantization (SQ) and 3) output probability recalculation. At the first step of the algorithm, the state probability is roughly estimated using the IDMM and SQ. The IDMM is an approximate computation of the multi-mixture probability density function, which realizes the fast probability estimation along with the SQ. The result of the rough estimation is used to select states with the high output probability. Then, the rigid probability calculation is carried out on the selected states. Our experiment on speaker independent continuous speech recognition showed that the proposed algorithm saves 81% of time for the output probability computation and 71% in the total speech recognition process, without degradation of the recognition rate.

To achieve robust speech recognition in noise, noise compensation techniques is necessary. The effectiveness of the parallel model combination technique (PMC) due to Gales (see Speech Communication, vo1.22, no.3, p.231-9, 1993) has been reported earlier. However, we found that the noise compensation mechanism using the basic PMC method didn't work well on the power parameters. We first show that there is a mismatch in the power parameters in the PMC method. Then, we propose a new method called IPPMC which calculates the power parameters independently of the other cepstral parameters. Finally we evaluate the IPPMC through the investigation of spectral envelopes, the comparison of the model distances to the target HMM trained by noisy speech and the experiment of the speaker-independent continuous speech recognition. The IPPMC improved the word accuracy from 82.1% obtained by the PMC to 85.5% at 10 dB SNR.