Perceptual Reversal over Time Course
(0) We present a computational model of perceptual reversal which alternates among two or more interpretations. Initially, the model represents the ambiguous atatf of the reversible picture such as facevasr. The internal state of the network evolves to settle into a stable slate which corresponds to one of alternatives. Top-down feedback in the model plays an important role that leads a whole system into its perceptual state over the time course. Also, topdown input from temporal associative memory interacts with bottom-up input, and gives a context to the network on how the bottom-up input can be inttrprettd at a given time. The model accounts for tht role of top-down knowledge in resolving perceptual ambiguity as well as reversibility from one state to the other.
KangWoo Lee Jianfeng Feng Hilary Buxton Soo-Young Lee
COGS, University of Sussex, Brighton, BN1 9QH, UK Brain Science Research Center, Korea Advanced Institute of Science and Technology, Tafjeon, Korea
国际会议
8th International Conference on Neural Information Processing(ICONIP 2001)(第八届国际神经信息处理大会)
上海
英文
64-70
2001-11-14（万方平台首次上网日期，不代表论文的发表时间）