Performance Improvement of Classifier by Combining Disjunctive Normal Form features

다국어 초록 (Multilingual Abstract)

This paper describes a visual object detection approach utilizing ensemble based machine learning.
Object detection methods employing 1D features have the benefit of fast calculation speed. However, for real image with complex background, detection accuracy and performance are degraded. In this paper, we propose an ensemble learning algorithm that combines a 1D feature classifier and 2D DNF (Disjunctive Normal Form) classifier to improve the object detection performance in a single input image. Also, to improve the computing efficiency and accuracy, we propose a feature selecting method to reduce the computing time and ensemble algorithm by combining the 1D features and 2D DNF features. In the verification experiments, we selected the Haar-like feature as the 1D image descriptor, and demonstrated the performance of the algorithm on a few datasets such as face and vehicle.

참고문헌 (Reference)

1 Oscar Martinez Mozos, "Supervised Learning of Places from Range Data using Adaboost" 1730-1735, 2005

2 Y. Bai, "Robust Tracking via Weakly Supervised Ranking SVM" 1854-1861, 2013

3 Paul Viola, "Rapid Object Detection using a Boosted Cascade of Simple Features" 1 : I-511-I-518, 2001

4 Robert T. Collins, "Online Selection of Discriminative Tracking Features" 27 (27): 1631-1643, 2005

5 D. Wang, "Online Object Tracking with Sparse Prototypes" 22 (22): 314-325, 2013

6 H. Grabner, "On-line boosting and vision" 1 : 260-267, 2006

7 Pedro F. Felzenszwalb, "Object Detection with Discriminatively Trained Part-Based Models" 32 (32): 9-2010, 2010

8 Timo Ojala, "Multiresolution Gray-Scale and Rotation Invariant Texture Classifiecation with Local Binary Patterns" 24 (24): 971-987, 2002

9 Kobi Levi, "Learning Object Detection from a Small Number of Examples: the Importance of Good Features" 2 : II-53-II-60, 2004

10 Takeshi Mita, "Joint Haar-like Features for Face Detection" 2 : 1619-1626, 2005

11 VOQUANG NHAT, "Illumination Invariant Object Tracking with Adaptive Sparse Representation" 제어·로봇·시스템학회 12 (12): 195-201, 2014

12 Rong Xiao, "Dynamic Cascades for Face Detection" 1-8, 2007

13 Zhu Teng, "Disjunctive Normal Form of Weak Classifier for Online Learning based Object Tracking" 138-146, 2013

14 Rainer Lienhart, "An Extended Set of Haar-like Features for Rapid Object Detection" 1 : I-900-I-903, 2002

15 김동욱, "A Fast and Accurate Face Tracking Scheme by using Depth Information in Addition to Texture Information" 대한전기학회 9 (9): 707-720, 2014