A fast method for training support vector machines with a very large set of linear features

J. Maydt1, R. Lienhart1
1Intel Laboratories, Intel Corporation, Santa Clara, CA, USA

Tóm tắt

Current systems for object detection often use support vector machines (SVM) as the basic classification algorithm. A rather common case is to compute a small set of linear features and then train the classifier on these features. We present a fast method to train and evaluate SVM with many linear features and show results for face detection using a set of 210400 features. The resulting classifier is both more accurate and faster than a classifier trained on raw pixel features, which total up only to 576 features in our case.

Từ khóa

#Support vector machines #Face detection #Object detection #Principal component analysis #Support vector machine classification #Classification algorithms #Kernel #Runtime #Fourier transforms #Wavelet domain

Tài liệu tham khảo

10.1109/ICIP.1999.819462 viola, 2001, Robust real-time object detection, Cambridge Research Laboratory Technical Report CRL 2001/01 serre, 2000, Feature Selection for Face Detection 10.1109/34.655647 0 chapelle, 2000, Choosing multiple parameters for support vector machines, Machine Learning 10.1023/A:1009715923555 maydt, 2001, People Detection in Digital Images A Component-Based Approach to Object Detection