An Adaptive Fuzzy Neural Network for Extracting Scene Image Parameters

As the most important objective parameters, reverberation time and clarity play significant roles in acoustic field characteristics evaluation of the hall. We can get it by measuring an actual room. In this paper, a new method is proposed based on adaptive fuzzy neural network to extract the reverberation time and clarity from a scene image. Finally the validity of the network is proved through the experiment results of network training on the test data. It provides a brand-new idea for virtual reality technique and sound quality evaluation of virtual environment with using this method.