Metaverse Space Ecological Scene Design Based on Multimedia Digital Technology

Yuan cosmos is a virtual world linked and created by scientific technological means, which mapped interacted with the real world, digital living space new social system. With increasing popularity of data acquisition production equipment, people are increasingly convenient to produce multimedia such as images, graphics, audio, video, animation, three-dimensional models. In addition rapid development technology itself, biological information related also greatly promotes emergence metauniverse. This paper aims study application ecological scene design metauniverse space, introduces concepts technology, expounds methods neural network algorithms, then takes simulation auditory system in interactive multisensory constituent elements an example. The mel-frequency cepstrum coefficient (MFCC) used simulate characteristics periphery (cochlea) perceptual end model. A variety bionic mechanisms model, designing connection mode neurons, learning state release effect, regeneration mechanism neurons. For verification performance speech sample database, including English words phrases, recorded content recognized model means recognition experimented. experimental results show that, terms phrase accuracy, DN-1 improves 2.59% DN-2 2.77% compared MFCC feature on basis mixed features. When only DBN features used, improvement rate developmental small.