혼합된 신호의 분리 학습을 위한 펄스 뉴런 경쟁학습 네트워크
Competitive Learning Pulsed Neural Network for a Separate Learning of the Mixed-signals
In this paper, the construction of the system identifying the environment around was studied by auditory information processing system using the neural network, until now a competitive learning neural network to determine the neuron of winner depending on the varying time of the input signal have been proposed. In case of appling competitive learning neural network using pulsed neurons to the real environment, there are disadvantages learning the mixed pattern in case the sound was generated from multiple sound sources, because the it quantized input pattern vectors immediately. In this paper, the new competitive learning network to learn by separating the signals generated from various sound sources taking the influence of reverb components etc is proposed by improving the competitive learning neural network using pulsed neurons. From the results of computer simulation, the vector quantization was confirmed by separating each sound in case of applying the new competitive learning network to the mixed signal lagging the starting point of sound.
- 원문이 없습니다.
- 원문이 없습니다.
유료 다운로드의 경우 해당 사이트의 정책에 따라 신규 회원가입, 로그인, 유료 구매 등이 필요할 수 있습니다. 해당 사이트에서 발생하는 귀하의 모든 정보활동은 NDSL의 서비스 정책과 무관합니다.
원문복사신청을 하시면, 일부 해외 인쇄학술지의 경우 외국학술지지원센터(FRIC)에서
무료 원문복사 서비스를 제공합니다.