Distinctive Image Features from Scale-Invariant Keypoints
International Journal of Computer Vision
A System Identification Approach for Video-based Face Recognition
ICPR '04 Proceedings of the Pattern Recognition, 17th International Conference on (ICPR'04) Volume 4 - Volume 04
Visual tracking and recognition using probabilistic appearance manifolds
Computer Vision and Image Understanding
Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories
CVPR '06 Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2
A dual coordinate descent method for large-scale linear SVM
Proceedings of the 25th international conference on Machine learning
Combining appearance and motion for face and gender recognition from videos
Pattern Recognition
Rotation Invariant Image Description with Local Binary Pattern Histogram Fourier Features
SCIA '09 Proceedings of the 16th Scandinavian Conference on Image Analysis
Evaluating Color Descriptors for Object and Scene Recognition
IEEE Transactions on Pattern Analysis and Machine Intelligence
Online learning from local features for video-based face recognition
Pattern Recognition
LIBSVM: A library for support vector machines
ACM Transactions on Intelligent Systems and Technology (TIST)
Video-based face recognition using probabilistic appearance manifolds
CVPR'03 Proceedings of the 2003 IEEE computer society conference on Computer vision and pattern recognition
Face detection based on multi-block LBP representation
ICB'07 Proceedings of the 2007 international conference on Advances in Biometrics
| Hi-index | 0.00 |
Video-based face recognition is a fundamental topic in image processing and video representation, and presents various challenges and opportunities. In this paper, we introduce an efficient patch-based bag of features (PBoF) method to video-based face recognition that plenty exploits the spatiotemporal information in videos, and does not make any assumptions about the pose, expressions or illumination of face. First, descriptors are used for feature extraction from patches, then with the quantization of a codebook, each descriptor is converted into code. Next, codes from each region are pooled together into a histogram. Finally, representation of the image is generated by concatenating the histograms from all regions, which is employed to do the categorization. In our experiments, 100% recognition rate is achieved on the Honda/UCSD database, which outperforms the state of the arts. And from the theoretical and experimental results, it can be derived that, when choosing a single descriptor and no prior knowledge about the data set and object is available, the dense SIFT with ScSPM is recommended. Experimental results demonstrate the effectiveness and flexibility of our proposed method.