Human Attribute Analysis Using a Top-View Camera Based on Two-Stage Classification
スポンサーリンク
概要
- 論文の詳細を見る
This paper presents a technique that analyzes pedestrians' attributes such as gender and bag-possession status from surveillance video. One of the technically challenging issues is that we use only top-view camera images to protect privacy. The shape features over the frames are extracted by bag-of-features (BoF) using histogram of oriented gradients (HoG) vectors. In order to enhance the classification accuracy, a two-staged classification framework is presented. Multiple classifiers are trained by changing the parameters in the first stage. The outputs from the first stage is further trained and classified in the second stage classifier. The experiments using 60-minute video captured at Haneda Airport, Japan, show that the accuracies for the gender classification and the bag-possession classification were 95.8% and 97.2%, respectively, which is a significant improvement from our previous work.
著者
-
Yamasaki Toshihiko
Department Of Electronic Engineering And Frontier Informatics The University Of Tokyo
-
MATSUNAMI Tomoaki
Department of Information and Communication Engineering, The University of Tokyo
-
CHEN Tuhan
School of Electrical and Computer Engineering, Cornell University
関連論文
- ウェアラブルビデオの環境移動変化シーン検出
- Robust Object-Based Watermarking Using Feature Matching
- ライフログの取得と検索のためのシステム構築
- LIFE LOG PLATFORM FOR CONTINUOUS AND DISCRETE RECORDING AND RETRIEVAL OF PERSONAL MEDIA(International Workshop on Advanced Image Technology 2007)
- Retrieval of Images Captured by Car Cameras Using Its Front and Side Views and GPS Data(Advanced Image Technology)
- A-16-21 Hierarchical Approach for Personal Life Media Retrieval
- Car Location Matching Using Front & Side-View and GPS Data(International Workshop on Advanced Image Technology 2006)
- D-12-77 Marker-less Pose Estimation for Fast Human Motion Clips
- A-16-14 Audio Segmentation using a Large Number of Microphones for Multimedia Retrieval in a Ubiquitous Environment
- A-15-15 Retrieval of Locomotion Patterns Using Sketch Based Spatial Queries
- A-20-12 Interactive Multimedia Retrieval Based on Illumination Changes in a Multi Camera Ubiquitous Environment
- 10-12 Video Summarization for a Large Number of Cameras Using Floor Sensors in a Ubiquitous Environment
- 9-2 Key Frame Extraction Based on Content and Context in Life Log Video
- 6-8 Surface Estimation Based on Model Segmentation
- Human Attribute Analysis Using a Top-View Camera Based on Two-Stage Classification
- SIFT-Based Non-blind Watermarking Robust to Non-linear Geometrical Distortions