Separating style and content with bilinear models

Separating style and content with bilinear models (2000)

Cached

Download Links

Other Repositories/Bibliography

DBLP

by Joshua B. Tenenbaum , William T. Freeman

Venue:	NEURAL COMPUTATION
Citations:	223 - 3 self

BibTeX

@ARTICLE{Tenenbaum00separatingstyle,
    author = {Joshua B. Tenenbaum and William T. Freeman},
    title = {Separating style and content with bilinear models},
    journal = {NEURAL COMPUTATION},
    year = {2000},
    pages = {1247--1283}
}

OpenURL

Abstract

PERCEPTUAL systems routinely separate content from style, classifying familiar words spoken in an unfamiliar accent, identifying a font or handwriting style across letters, or recognizing a familiar face or object seen under unfamiliar viewing conditions. Yet a general and tractable computational model of this ability to untangle the underlying factors of perceptual observations remains elusive. Existing factor models are either insufficiently rich to capture the complex interactions of perceptually meaningful factors such as phoneme and speaker accent or letter and font, or do not allow efficient learning algorithms. Here we show how perceptual systems may learn to solve these crucial tasks using surprisingly simple bilinear models. We report promising results in three realistic perceptual domains: spoken vowel classification with a benchmark multi-speaker database, extrapolation of fonts to unseen letters, and translation of faces to novel illuminants.

Keyphrases

bilinear model perceptual system complex interaction perceptual observation benchmark multi-speaker database simple bilinear model unseen letter tractable computational model familiar face separate content familiar word vowel classification realistic perceptual domain speaker accent factor model meaningful factor underlying factor crucial task unfamiliar accent