D-12-61 The pedestrian Tracking Based on an On-line Boosting Method
スポンサーリンク
概要
- 論文の詳細を見る
In actual surveillance conditions, there exist a lot of uncertainties in pedestrian movements. These movements may disturb most tracking algorithms and result in tracking failure. In this paper, a new pedestrian tracking system is proposed, in which an online-boosting method is embedded into the pedestrian tracking of object model. Online-boosting [1] methods have become very popular in computer vision, and achieved impressive performance in detection and recognition tasks. However, the on-line adaption has to face one key problem: an error which finally leads to tracking failure (drifting) may be accumulated in each update of the tracking. In order to avoid the drifting problem, MIL (Multiple Instance Learning) method is proposed, instead of traditional supervised learning, to lead to a more robust tracker with well performance.
- 2011-02-28
著者
-
Nagahashi Hiroshi
Imaging Science And Engineering Laboratory Tokyo Institute Of Technology
-
Wang Qian
Interdisciplinary Graduate School Of Science And Engineering Tokyo Institute Of Technology
-
Aoki Kota
Imaging Science and Engineering Laboratory, Tokyo Institute of Technology
-
Aoki Kota
Imaging Science And Engineering Laboratory Tokyo Institute Of Technology
関連論文
- Structural Evolution of Neural Networks Having Arbitrary Connections by a Genetic Method
- A New Method for Smooth Interpolation without Twist Constraints
- DISTANCE MEASUREMENT OF A REAL-WORLD ENVIRONMENT USING AN ACTIVE CAMERA SYSTEM(International Workshop on Advanced Image Technology 2006)
- Interpolation of CT Slices for Laser Stereolithography
- Orientable Closed Surface Construction from Volume Data
- A Method for Watermarking to Bezier Polynomial Surface Models (Computer Graphics)
- Texture Classification Using Hierarchical Linear Discriminant Space(Image Recognition, Computer Vision)
- Parallel Computation of Parametric Piecewise Modeling Method
- Texture-based Real-time Vision System for Detection of Automobile Cylinder Sleeves
- D-12-43 Photograph Based Pair-matching Recognition of Human Faces
- D-12-61 The pedestrian Tracking Based on an On-line Boosting Method
- Verification of physical properties of materials modeled with mass-spring systems (マルチメディア・仮想環境基礎)
- Verification of physical properties of materials modeled with mass-spring systems (画像工学)
- D-16-5 Multifractal Analysis for HCC Biopsy Images
- Classification of Prostate Histopathology Images Based on Multifractal Analysis
- D-12-22 A Simple Approach Tackling Illumination Problem of Face Detection
- Study on Illumination Insensitive Face Detection Based on Normalization Techniques