Abstract

Object identification (OID) is specialized recognition where the category is known (e.g. cars) and the algorithm recognizes an object’s exact identity (e.g. Bob’s BMW). Two special challenges characterize OID. (1) Inter-class variation is often small (many cars look alike) and may be dwarfed by illumination or pose changes. (2) There may be many classes but few or just one positive “training ” examples per class. Due to (1), a solution must locate possibly subtle objectspecific salient features (a door handle) while avoiding distracting ones (a specular highlight). However, (2) rules out direct techniques of feature selection. We describe an online algorithm that takes one model image from a known category and builds an efficient “same ” vs. “different ” classification cascade by predicting the most discriminative feature set for that object. Our method not only estimates the saliency and scoring function for each candidate feature, but also models the dependency between features, building an ordered feature sequence unique to a specific model image, maximizing cumulative information content. Learned stopping thresholds make the classifier very efficient. To make this possible, category-specific characteristics are learned automatically in an off-line training procedure from labeled

Citations

3104	Distinctive image features from scale invariant keypoints - Lowe
944	J: General Linear Models - McCullagh, Nelder - 1989
515	K: Irrelevant features and the subset selection problem - GH, Kohavi, et al.
137	A Bayesian approach to unsupervised one-shot learning of object categories - Fei-Fei, Fergus, et al.
91	R: Why faces are and are not special: An effect of expertise - Carey, Diamond - 1986
75	Object recognition with informative features and linear classification - Vidal-Naquet, Ullman
73	Learning from one example through shared densities on transforms - Miller, Matsakis, et al. - 2000
56	FFA: A flexible fusiform area for subordinate-level visual processing automatized by expertise. Nature Neuroscience 2000;3:764–9 - MJ, Gauthier
39	The CSU face identification evaluation system: Its purpose, features, and structure - Bolme, Beveridge, et al. - 2003
23	Unsupervised learning of models for recognition. ECCV - Weber, Welling, et al. - 2000
13	Learning hyper-features for visual identification - Ferencz, Learned-Miller, et al. - 2004
10	Learning to locate informative features for visual identification - Ferencz, Learned-Miller, et al. - 2007
4	A two-variate gamma type distribution - Kibble - 1941
1	Learning hyper-features for visual identification - Anonymous