Transfer forest based on covariate shift

Random Forest, a multi-class classifier based on statistical learning, is widely used in applications because of its high generalization performance due to randomness. However, in applications such as object detection, disparities in the distributions of the training and test samples from the target scene are often inevitable, resulting in degraded performance. In this case, the training samples need to be reacquired for the target scene, typically at a very high human acquisition cost. To solve this problem, transfer learning has been proposed. In this paper, we present data-level transfer learning for a Random Forest using covariate shift. Experimental results show that the proposed method, called Transfer Forest, can adapt to the target domain by transferring training samples from an auxiliary domain.

[1]  Yamauchi Yuji,et al.  Hybrid Transfer Learning for Efficient Learning in Object Detection , 2013 .

[2]  Klaus-Robert Müller,et al.  Covariate Shift Adaptation by Importance Weighted Cross Validation , 2007, J. Mach. Learn. Res..

[3]  Meng Wang,et al.  Transferring a generic pedestrian detector towards specific scenes , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[4]  Thomas G. Dietterich,et al.  To transfer or not to transfer , 2005, NIPS 2005.

[5]  Tong Zhang,et al.  A Framework for Learning Predictive Structures from Multiple Tasks and Unlabeled Data , 2005, J. Mach. Learn. Res..

[6]  Pietro Perona,et al.  One-shot learning of object categories , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[7]  Thomas G. Dietterich,et al.  Transfer Learning with an Ensemble of Background Tasks , 2005, NIPS 2005.

[8]  Tom Heskes,et al.  Task Clustering and Gating for Bayesian Multitask Learning , 2003, J. Mach. Learn. Res..

[9]  Qiang Yang,et al.  Boosting for transfer learning , 2007, ICML '07.

[10]  Dariu Gavrila,et al.  Monocular Pedestrian Detection: Survey and Experiments , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[11]  Roberto Cipolla,et al.  Semantic texton forests for image categorization and segmentation , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[12]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[13]  Ali Farhadi,et al.  Learning to Recognize Activities from the Wrong View Point , 2008, ECCV.

[14]  Antonio Criminisi,et al.  Decision Forests: A Unified Framework for Classification, Regression, Density Estimation, Manifold Learning and Semi-Supervised Learning , 2012, Found. Trends Comput. Graph. Vis..

[15]  Vincent Lepetit,et al.  Randomized trees for real-time keypoint recognition , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[16]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[17]  Sebastian Thrun,et al.  Is Learning The n-th Thing Any Easier Than Learning The First? , 1995, NIPS.

[18]  Alan Fern,et al.  Multi-task reinforcement learning: a hierarchical Bayesian approach , 2007, ICML '07.

[19]  Qingming Huang,et al.  Transferring Boosted Detectors Towards Viewpoint and Scene Adaptiveness , 2011, IEEE Transactions on Image Processing.

[20]  Hironobu Fujiyoshi,et al.  Human detection for multiple pose by boosted randomized trees , 2011, The First Asian Conference on Pattern Recognition.