Visual Object Identification

	Building Classification Cascades for Visual Identification from One Example Andras Ferencz UC Berkeley & Mobileye Vision Technologies Erik G. Learned-Miller UC Berkeley & UMass Amherst Jitendra Malik UC Berkeley Vidit Jain UMass Amherst
Summary	In this project, we attempt to solve the problem of object identification (OID), which is specialized recognition where the category is known (for example cars or faces) and the algorithm recognizes an object's exact identity (such as Bob's BMW). For example, we might be given images of cars like these: and be asked to find which of the 4 cars on right are the same as either of the 2 on the left. Two special challenges characterize OID: Inter-class variation is often small (many cars look alike) and may be dwarfed by illumination or pose changes. There may be many classes but few or just one positive ``training'' examples per class. Due to (1), a solution must locate possibly subtle object-specific salient features (a door handle) while avoiding distracting ones (a specular highlights). However, (2) rules out direct techniques of feature selection. We describe an on-line algorithm that takes one query 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 query 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 image pairs of the category, without prior knowledge about the category. In functional notation, our algorithm has 3 steps: (Off-line) trainer T_id: category training images --> identifier generator H_id (On-line) identifier generator H_id: object image --> classifier C_id (On-line) classifier C_id: test image --> { same, different } Where step 1 learns category specific characteristics, while step 2 creates an object specific classifier.
Publications	Andras Ferencz, Erik Learned-Miller, Jitendra Malik. Learning to Locate Informative Features for Visual Identification. Draft: submitted to IJCV Special Issue: Learning for Vision, 2007. PDF Vidit Jain, Andras Ferencz, Erik Learned-Miller. Discriminative Training of Hyper-feature Models for Object Identification. In Proceedings of British Machine Vision Conference, 2006. PDF [Project Page @ UMass Amherst] Andras Ferencz, Erik Learned-Miller, Jitendra Malik. Building a Classification Cascade for Visual Identification from One Example. ICCV 2005. PDF Andras Ferencz, Erik Learned-Miller, Jitendra Malik. Learning Hyper-Features for Visual Identification. Neural Information Processing Systems, 2004. PDF
Presentations	Dissertation Talk. March 2005. PDF NIPS*05 Interclass Transfer Workshop. [PDF]
ExampleResults	Car and face classification examples (with the matched patches marked).
Downloads	Our car data set and an implementation of our system is available for download here.
Martians	The learning problem that must be solved in visual identification (steps 1 & 2 above) is best demonstrated with an example: If we knew nothing about Martians, finding Bob after having seen only a single image of him before, would be impossible. However, with a training set of other Martians, even if Bob is not in this set, the problem becomes tractable.