DocumentCode :
1621485
Title :
Hierachical learning of visual invariances via spatio-temporal constraints
Author :
Stone, J.V.
Author_Institution :
Sussex Univ., Brighton, UK
fYear :
1995
Firstpage :
110
Lastpage :
115
Abstract :
An unsupervised learning algorithm is presented for the acquisition of low-level vision tasks, such as the extraction of surface depth. A key assumption is that perceptually salient parameters in the image of a moving object vary smoothly over time. This assumption is consistent with a learning rule which maximises the long-term variance of a unit´s outputs, whilst simultaneously minimising its short-term variance. This learning rule involves a linear combination of anti-Hebbian and Hebbian weight changes, over short and long time scales respectively. It can be shown that the learning rule maximises the mutual information between a unit´s output and the short-term exponentially weighted mean of that unit´s output. The model is demonstrated on a hyper-acuity task: estimating sub-pixel stereo disparity from a temporal sequence of stereograms. The model can discriminate disparities as small as one-tenth of the width of an input receptor, which compares with human performance on stereo hyper-acuity tasks. Furthermore, the algorithm generalises, without additional learning, to previously unseen image sequences. The results of applying the learning method independently to successive layers of a hierarchical network are presented
Keywords :
Hebbian learning; computer vision; generalisation (artificial intelligence); hierarchical systems; image sequences; invariance; neural nets; spatial reasoning; stereo image processing; temporal reasoning; unsupervised learning; Hebbian weight changes; antiHebbian weight changes; generalization; hierachical learning; hierarchical network; hyper-acuity task; input receptor width; low-level vision task acquisition; moving object image; mutual information; short-term exponentially weighted mean; smoothly varying perceptually salient parameters; spatiotemporal constraints; stereogram temporal sequence; sub-pixel stereo disparity estimation; surface depth extraction; unit output long-term variance maximization; unit output short-term variance minimization; unseen image sequences; unsupervised learning algorithm; visual invariances;
fLanguage :
English
Publisher :
iet
Conference_Titel :
Artificial Neural Networks, 1995., Fourth International Conference on
Conference_Location :
Cambridge
Print_ISBN :
0-85296-641-5
Type :
conf
DOI :
10.1049/cp:19950538
Filename :
497800
Link To Document :
بازگشت