V. P. Kumar and T. Poggio. "Learning-based approach to real time tracking and analysis of faces," in Proc. of AFGR, 2000.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....with just 11 of false positives. This is a data analysis approach that will help to many skin detection systems. 1 Introduction Skin detection is a very important step in many vision systems like gestures, hand tracking, video indexing, region of interests, face detection, etc. see e.g. [2, 4, 6, 10, 14 17] to name just a few) Pixel based skin detection can narrow the search space prior to high level layers. However, this is not an easy task. Skin values can vary with ambient light, such as colour lamps acting as filters, specularities, shadows, daylight, etc. Moreover, di#erent cameras returns ....

V. P. Kumar, T. Poggio. Learning-based approach to real time tracking and analysis of faces. Automatic Face and Gesture Recognition, pp. 96-101, 2000.

....compute the matching parameters with considerably less computations and using only view based representations would make these models useful in real time applications. The motivation for this work comes from the use of a learning based approach in real time analysis of mouths (Kumar and Poggio [7]) in which it was shown that a regression function can be learnt from a Haar wavelet based input representation of mouths to hand labeled parameters denoting openness and smile. Therefore, it points to the possibility that learning may be a way for directly estimating the matching parameters of ....

....[4] Thus each image is represented as a six dimensional vector, which form the outputs for the learning problem. Each of the 2066 images is subject to the Haar wavelet transform and feature selection involving selection of those Haar coefficients with the highest variance (See Kumar and Poggio [7]) We select 12 coefficients with the highest variance which form the inputs for the learning problem. 3.2 Training the SVM based regression In this section, we sketch the ideas behind using SVM for learning regression functions (a more detailed description can be found in Golowich, et al. 8] ....

V. Kumar and T. Poggio. Learning-Based Approach to Real Time Tracking and Analysis of Faces. In Proceedings of the Fourth International Conference on Automatic Face and Gesture Recognition, pages 96 101, Grenoble, letante, 2000.

....compute the matching parameters with considerably less computations and using only view based representations would make these models useful in real time applications. The motivation for this work comes from the use of a learning based approach in real time analysis of mouths (Kumar and Poggio [7]) in which it was shown that a regression function can be learnt from a Haar wavelet based input representation of mouths to hand labeled parameters denoting openness and smile. Therefore, it points to the possibility that learning may be a way for directly estimating the matching parameters of ....

....Thus each image is represented as a six dimensional vector, which form the outputs for the learning problem. 3 . Each of the 2066 images is subject to the Haar wavelet transform and feature selection involving selection of those Haar coe#cients with the highest variance (See Kumar and Poggio [7]) We select 12 coe#cients with the highest variance which form the inputs for the learning problem. 3.2 Training the SVM basI regres7I6 In this section, we sketch the ideas behind using SVM for learning regression functions (a more detailed description can be found in Golowich, et al. 8] and ....

V. Kumar and T. Poggio. Learning-Based Approach to Real Time Tracking and Analysis of Faces. In Proceeding of the Fourth International Conference on Automatic Face and Gesture RecogF-S3V , pages 96--101, Grenoble, France, 2000.

....a sufficient degree of reliability, depending on the specific application and security level. The development of automatic visual surveillance systems can now leverage techniques for detecting and recognizing people that have been developed recently: pedestrian detection [1, 2] face detection [3, 4, 5], face recognition [6, 7] and motion detection [8, 9] In general, the unconstrained task of people recognition still presents a number of difficult challenges due to the similarity of people images, pose variations, change of clothes, different illumination and background conditions. In our ....

V. P. Kumar and T. Poggio. Learning-based approach to real time tracking and analysis of faces. Proc. of AFGR, 2000.

....to be made. A central aim of further work is to link the existing algorithm to a robust method for automatic tracking. Such methods have been proposed in the literature (e.g. 4, 23] An interesting alternative strategy is motivated by a recent results on the recognition of facial expressions [15]. In this system, the mapping between the outputs of a Gabor filter bank onto low dimensional weights of a morphable model for stationary images was learned using support vector regression. Using such low dimensional weights instead of tracked feature position as input for our algorithm would ....

V. Kumar and T. Poggio. Learning-based approach to realtime tracking analysis of faces. Technical Report 1672, Massachusetts Institute of Technology, Cambridge, MA, 1998.

No context found.

V. P. Kumar and T. Poggio. "Learning-based approach to real time tracking and analysis of faces," in Proc. of AFGR, 2000.

No context found.

V. P. Kumar and T. Poggio. "Learning-based approach to real time tracking and analysis of faces," in Proc. of AFGR, 2000.

No context found.

Kumar, V. P., Poggio, T.: Learning-based approach to real time tracking and analysis of faces. Proc. 4th IEEE Int. Conf. Automatic Face and Gesture Recognition. Grenoble, France (March 2000) 96--101.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC