Tracking a detected face with dynamic programming

Yao, Zhengrong; Li, Haibo

doi:10.1016/j.imavis.2005.09.007

Cited by 18 publications

(5 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, for the sequences of up, down, left, and right facial movements, such as #7 and #9, our system generally showed a higher tracking rate. In particular, our system was much better than method [11] in which the front of the face does not appear during the given time and the face is extremely rotated (sequences #4 and #6) when the sequence involved a moving face(sequences #1 and #3) as shown in fig. 11.…”

Section: Comparison With Related Workmentioning

confidence: 89%

“…Table 1. The comparison of the obtained results using our system and those obtained using method [11] Method [11] Our When we apply our method to an image(#8), the track rate is less than that of method [11]. For face candidates the face image is not detected accurately because the face is hidden by a shadow.…”

Section: Comparison With Related Workmentioning

confidence: 93%

“…To compare our tracking system with related works, we selected a method from among the various tracking methods. This method [11] uses Viola-Jones's basic features and Lienhart's [12] extended Haar-like features in order to increase the detection rate of various facial poses. It creates a one-dimensional deformable face graph after subdividing the eyes and mouth in the detected face.…”

Section: Comparison With Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Real-Time Face Tracking System Using Adaptive Face Detector and Kalman Filter

Kim

Kang

Eom

et al. 2007

Human-Computer Interaction. HCI Intelligent Multimodal Interaction Environments

View full text Add to dashboard Cite

In this paper, we propose a real-time face tracking system using adaptive face detector and the Kalman filter. Basically, the features used for face detection are five types of simple Haar-like features. To only extract the more significant features from these features, we employ principal component analysis (PCA). The extracted features are used for a learning vector of the support vector machine (SVM), which classifies the faces and non-faces. The face detector locates faces from the face candidates separated from the background by using real-time updated skin color information. We trace the moving faces with the Kalman filter, which uses the static information of the detected faces and the dynamic information of changes between previous and current frames. In this experiment, the proposed system showed an average tracking rate of 97.3% and a frame rate of 23.5 frames per s, which can be adapted into a realtime tracking system.

show abstract

Section: Comparison With Related Workmentioning

confidence: 89%

Section: Comparison With Related Workmentioning

confidence: 93%

Section: Comparison With Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Real-Time Face Tracking System Using Adaptive Face Detector and Kalman Filter

Kim

Kang

Eom

et al. 2007

Human-Computer Interaction. HCI Intelligent Multimodal Interaction Environments

View full text Add to dashboard Cite

show abstract

“…Face analysis is a method used in several applications such as cell phone security systems, to detect stress levels, in gender recognition, etc. [1] [2] [3]; Another topic of interest is undoubtedly the follow-up of the face and its analysis in video sequences [4] [5]; and it is because there is more information in a temporary space than if only one image is analyzed; since, without a doubt, the dependence between one image and another on a video sequence reveals information of interest; However, the problem is complicated, if the video is of poor quality as are internet videos.…”

Section: Introductionmentioning

confidence: 99%

Video Analysis with Faces using Harris Detector and Correlation

Herrera¹,

Gallegos-Funes²,

Martínez³

2019

IJACSA

View full text Add to dashboard Cite

A procedure is presented to detect changes in a video sequence based on the Viola & Jones Method to obtain images of faces of persons; videos are taken of the web. The software allows to obtain images or frames separated from nose, mouth, and eyes but the case of the eyes is taken as an example. Change detection is done by using correlation and the Harris detector. The correlation results allow us to recognize changes in position in the person or movement of the camera if the individual remains fixed and vice versa. It is possible to analyze a part of the face thanks to the Harris detector; It is possible with detected reference points to recognize very small changes in the video sequence of a particular organ; such as the human eye, even when the image is of poor quality as is the case of videos downloaded from the Internet or taken with a low-resolution camera.

show abstract

“…Numerous facial tracking methods have been proposed. These include methods based on intensity gradients [9], graph matching (deformable templates) [10], and contour tracking (snakes) [11], all of which provide promising results but require numerous calculations or extensive statistical datasets. Elaborate systems implementing Kalman filtering and kmean clustering [12] or adaptive hidden Markov model (HMM) classification [13], provide increased accuracy at the expense of still more additional computations.…”

Section: Introductionmentioning

confidence: 99%

Face and lip tracking in unconstrained imagery for improved automatic speech recognition

Crow

Zhang

2009

SPIE Proceedings

View full text Add to dashboard Cite

When combined with acoustical speech information, visual speech information (lip movement) significantly improves Automatic Speech Recognition (ASR) in acoustically noisy environments. Previous research has demonstrated that visual modality is a viable tool for identifying speech. However, the visual information has yet to become utilized in mainstream ASR systems due to the difficulty in accurately tracking lips in real-world conditions. This paper presents our current progress in tracking face and lips in visually challenging environments. Findings suggest the mean shift algorithm performs poorly for small regions, in this case the lips, but it achieves near 80% accuracy for facial tracking.

show abstract

Tracking a detected face with dynamic programming

Cited by 18 publications

References 11 publications

Real-Time Face Tracking System Using Adaptive Face Detector and Kalman Filter

Real-Time Face Tracking System Using Adaptive Face Detector and Kalman Filter

Video Analysis with Faces using Harris Detector and Correlation

Face and lip tracking in unconstrained imagery for improved automatic speech recognition

Contact Info

Product

Resources

About