Learning a Family of Detectors via Multiplicative Kernels

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 33; no. 3; pp. 514 - 530
• Main Authors: Quan Yuan, Thangali, Ashwin, Ablavsky, Vitaly, Sclaroff, Stan
• Format: Journal Article
• Language: English
• Published: Los Alamitos, CA IEEE 01.03.2011; IEEE Computer Society; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Applied sciences; Artificial Intelligence; Classification; Computer science; control theory; systems; Computer Simulation; Detectors; Exact sciences and technology; Humans; Image Enhancement - methods; Image Interpretation, Computer-Assisted - instrumentation; Imaging, Three-Dimensional - instrumentation; Kernel; kernel methods; Learning; Markov Chains; Mathematical models; Motion; Motor Vehicles; Numerical Analysis, Computer-Assisted - instrumentation; Object detection; Object recognition; Object segmentation; object tracking; Pattern Recognition, Automated - methods; Pattern recognition. Digital image processing. Computational geometry; Phantoms, Imaging; pose estimation; Reproducibility of Results; Sensitivity and Specificity; Subtraction Technique - instrumentation; Support vector machine classification; Support vector machines; Tasks; Tracking; Training; Vehicle detection; Vehicles; Video sequences; Computer vision; Target tracking; Tracking; Segmentation; Mask; Pattern recognition; Object recognition; Modeling; Kernel method; Posture; Kernel function; Surveillance; Image sequence; Scene analysis; Facies; Object detection; pose estimation; Vector support machine; kernel methods; object tracking; Moving body
• ISSN: 0162-8828; 1939-3539; 1939-3539
• DOI: 10.1109/TPAMI.2010.117

Object detection is challenging when the object class exhibits large within-class variations. In this work, we show that foreground-background classification (detection) and within-class classification of the foreground class (pose estimation) can be jointly learned in a multiplicative form of two kernel functions. Model training is accomplished via standard SVM learning. When the foreground object masks are provided in training, the detectors can also produce object segmentations. A tracking-by-detection framework to recover foreground state in video sequences is also proposed with our model. The advantages of our method are demonstrated on tasks of object detection, view angle estimation, and tracking. Our approach compares favorably to existing methods on hand and vehicle detection tasks. Quantitative tracking results are given on sequences of moving vehicles and human faces.