Learn More
In many applications, Chinese information is very often provided in the form of phonetic symbol sequences, and it is desired to decode such sequences into the corresponding Chinese character sequences (sentences) as the output. Phonetic input of Chinese characters into computers is a typical example. The problem is due primarily to the high degree of(More)
AhtractThis paper describes the first successfully implemented real-time Mandarin dictation machine developed in the world which recognizes Mandarin speech with very large vocabulary and almost unlimited texts for the input of Chinese characters into computers. Considering the special characteristics of the Chinese language, syllables are chosen as the(More)
In this paper, a method with sentence-wide optimization consideration is proposed to generate a Mandarin sentence's pitch-contour. The developed model is called the sentence pitch-contour HMM (SPC-HMM) due to its use of VQ (vector quantization) and HMM (hidden Markov model). To construct an SPC-HMM, the pitch-contours of the syllables from each training(More)
In this paper, HNM (harmonic plus noise model) is enhanced and used to design a scheme for synthesizing Mandarin singing voice. Enhancements made include synthesizing signals with higher fluency level and keeping the timbre of synthetic singing voices consistent. In terms of the signal synthesis equations rewritten here, a Mandarin singing voice synthesis(More)
This paper aims to report on the technical development and the realization of the world first robot theatre performance with a cast comprising two biped androids and two twin-wheeled two-armed humanoid robots. Each of the biped androids has a head with human-face, capable of showing multiple facial expressions, and a pair of 7 DOF arms with a multi-jointed(More)
To reduce the call blocking probability (CBP) and call dropping probability (CDP) of real-time and non-real-time traffic in wireless multimedia networks, it is frequent to employ resource reservation to achieve the above-mentioned goal as well as to avoid long latency of path rebuilding. However, pure resource reservation may lead to inefficient resource(More)
除了 LPC 之外,過去也有幾個以倒頻譜(cepstrum)為基礎的頻譜包絡估計方法被提 出,最簡單的一個是倒頻譜平滑法[1],此法只保留倒頻譜係數的前幾個,而把後面的 係數全部砍除(即令為 0 值),再作離散傅利葉轉換(discrete Fourier transform , DFT),就 可得到平滑的頻譜曲線,如圖 1 裡下方的那一條平滑曲線,很明顯地這樣的一條頻譜曲 線並不是頻譜包絡,因為它走在原始 DFT 頻譜的波峰與波谷之間,而不是沿著波峰行 走。因此,Imai 和 Abe 兩人提出一個以倒頻譜為基礎再作改進的方法[3, 4] ,稱為 true envelope 估計法,然而此法的計算量很大而缺乏效率。另外,Galas 和 Rodet 兩人提出 以離散倒頻譜(discrete cepstrum(More)
In this paper, an approach to compress Chinese text is proposed. This approach first extends the alphabet to include those Chinese characters in Big-5 code. Then, an adaptive Markov model is used to model the contextual dependency, and arithmetic coding is used to encode the data more compactly. In the case of large alphabet size, a practical implementation(More)
In synthetic Mandarin speech, discontinuity of formant traces at syllable boundaries is a key factor that lowers the fluency level. Therefore, we study an acoustic and articulatory knowledge integrated method to solve this discontinuity problem. First, representative trisyllable contexts are selected and their signals are recorded. The signal of the middle(More)