| 1. | Main ideas as follows : research of audio digital watermark algorithm based on fast fourier transforms and psychoacoustic auditory model
主要工作有:基于快速傅立叶变换与声学模型的音频数字水印算法研究。 |
| 2. | In this dissertation , psychoacoustic models filter banks quantization and bit allocation predicate coding and temporary noise shaping is analyzed
本文首先对心理声学模型、滤波器组、量化和比特分配、预测编码、时域噪声成型等技术进行研究。 |
| 3. | A parametric psychoacoustic model based on critical band rate , level - dependent spreading function , non - linear superposition and forward masking was proposed
摘要提出了基于临界频带率、依赖于强度的扩展函数、非线性迭加和前向掩蔽效应的参数心理声学模型。 |
| 4. | Journal of chinese information processing , 2004 , 18 : 78 - 84 . 74 zhu xiaoyan , wang yu , xu wei . speech recognition model based on recurrent neural networks
中文语音识别在经典的隐马尔科夫模型基础上又提出了一系列改进模型，如分割模型聚类模型双语声学模型等。 |
| 5. | The experiments verify that with these new technologies , the amount of sinusoids can be greatly reduced and the tracking efficiency can be greatly improved
实验证明，将这两种方法与参数心理声学模型相结合，可以有效地减少所需的正弦波数量，大大提高正弦波跟踪的效率。 |
| 6. | Furthermore , on the analysis of the psychoacoustic models , some simplification can be made in the psychoacoustic models . the purpose of this simplification is to reduce the bit rate of audio stream
此外，基于对心理声学模型的分析，对心理声学模型作若干近似，简化的目的是为了用较少的比特编码音频数据。 |
| 7. | The newly proposed method makes good use of the psychoacoustic model . and the perceptual information contained in the synthesized signal is increased in a furthest way during the modeling procedure
感知梯度正弦建模比较好地利用了心理声学模型，在建模过程中最大限度地增加合成信号的感知信息，提高了建模效率。 |
| 8. | The frequency analysis of the psychoacoustic model can give the inaudible , redundan elements of audio . in this paper we employ it to guarantee the inaudibility of the embedded watermark
心理声学模型的时频域分析主要用于音频数据的压缩编码，这里主要考虑将其用于音频水印算法中去，以便从理论的角度来考虑嵌入水印的不可感知性。 |
| 9. | Second , we analysis the character of the human auditory system , especially the psychoacoustic model . then we analyze the principle of the audio part of mpeg - 1 which apply the psychoacoustic model to compressing the audio data
本文总结了人类听觉系统的特点，特别是人类听觉的感知特性，分析了mpeg标准音频部分所使用的心理声学模型。 |
| 10. | This paper also does some primary research for model experiment if this prediction way , model experiment and fact measure can be fell together , then some factors which is difficult to consider will be corrected , by this way the result will be more precise
本文还对声学模型实验进行了一些初步的探索，若能将本文的预测方法与模型实验和实测结合起来，对一些难以考虑的因素进行修正，必将得到更加准确的结果。 |
