Talking Robot and the Autonomous Acquisition of Vocalization and Singing Skill
Talking Robot and the Autonomous Acquisition of Vocalization and Singing Skill
复制标题
说话机器人与发声和歌唱技能的自主习得
DOI：
10.5772/4761
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
H. Sawada
中科院分区：
文献类型：
--
作者：
H. Sawada
语音是人类交流的主要媒介。它不仅用于简单的日常交流，也用于逻辑讨论和情感表达。不同的声音是在听觉系统的反馈控制机制下，由发声器官的复杂运动产生的。声音和人的发声机制一直是许多研究者关注的研究课题，而计算机化的声音产生和识别已经成为近年来柔性人机界面研究的关键技术。在声音产生的研究中，已经报道了各种方法和技术。算法合成已经取代了模拟电路合成，成为广泛应用的技术[2]，[3]。声音采样方法和基于物理模型的合成是典型的技术，它们有望提供不同类型的真实声音[4]。除了这些算法合成技术之外，使用模仿人类发声机制的语音或声乐模型的机械方法将是一个有价值和值得注意的目标。已经报道了几种人类发声系统的机械结构来实现类似人类的语言。然而，在大多数研究[2]，[5]，[6]中，人类发声系统的机械复制主要是参考x射线图像和FEM分析，迄今尚未考虑自然发声控制方法的自适应获取。事实上，由于流体动力学的非线性因素尚未克服，对发声器官行为的研究还不够充分，机械系统的控制往往是建立的难点。为了实现模仿人类发声的说话机器人[7]-[11]，作者开发了一种机械语音生成系统及其控制技巧的自适应学习。基频和频谱包络决定了声音的主要特性。前者是由振动物体产生的声源声的特性，后者是由谐振效应的工作运作的。在发声过程中，声带的振动产生一个声源，然后声波被引导到声道，声道作为一个过滤器来确定频谱包络。
Voice is used as primary media in the human communication. It is employed not only in simple daily communication, but also for the logical discussions and the expression of emotion and feelings. Different vocal sounds are generated by the complex movements of vocal organs under the feedback control mechanisms using an auditory system. Vocal sounds and human vocalization mechanisms have been the attractive researching subjects for many researchers so far [1],[2], and computerized voice production and recognition have become the essential technologies in the recent developments of flexible human-machine interface studies. Various ways and techniques have been reported in the researches of sound production. Algorithmic syntheses have taken the place of analogue circuit syntheses and became widely used techniques [2],[3]. Sound sampling methods and physical model based syntheses are typical techniques, which are expected to provide different types of realistic vocal sounds [4]. In addition to these algorithmic synthesis techniques, a mechanical approach using a phonetic or vocal model imitating the human vocalization mechanism would be a valuable and notable objective. Several mechanical constructions of a human vocal system to realize human-like speech have been reported. In most of the researches [2],[5],[6], however, the mechanical reproductions of the human vocal system were mainly directed by referring to X-ray images and FEM analysis, and the adaptive acquisition of control methods for natural vocalization have not been considered so far. In fact, since the behaviours of vocal organs have not been sufficiently investigated due to the nonlinear factors of fluid dynamics yet to be overcome, the control of mechanical system has often the difficulties to be established. The author has been developing a mechanical voice generation system together with its adaptive learning of the control skill for the realization of a talking robot which imitates human vocalization [7]-[11]. The fundamental frequency and the spectrum envelope determine the principal characteristics of a sound. The former is the characteristic of a source sound generated by a vibrating object, and the latter is operated by the work of the resonance effects. In vocalization, the vibration of vocal cords generates a source sound, and then the sound wave is led to a vocal tract, which works as a filter to determine the spectrum envelope.
说话机器人及其自主语音采集分析
DOI：
--
发表时间：
2006
期刊：
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS2006)
影响因子：
--
作者：
Tan;J.K.;Ishikawa;S;Mitsuhiro Nakamura and Hideyuki Sawada
通讯作者：
Mitsuhiro Nakamura and Hideyuki Sawada
机械发声系统及其歌唱性能
DOI：
--
发表时间：
2004
期刊：
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS2004), <Hyper Human Tech Award受賞> 2004
影响因子：
--
作者：
Hideyuki Sawada;Mitsuhiro Nakamura
通讯作者：
Mitsuhiro Nakamura