Locating human faces in newspaper photographs

Govindaraju, Venu; Sher, David B.; Srihari, Rohini K.; Srihari, Sargur N.

doi:10.1109/cvpr.1989.37900

Cited by 48 publications

(18 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For example, the edge map of the image [12], [26], [20], [7], [40], [14] is often considered as the basic image representation model for general object recognition and, in particular, for face recognition [22], [47], [17], [3]. Other examples of image representations will be considered below later.…”

Section: Introductionmentioning

confidence: 99%

Face recognition: the problem of compensating for changes in illumination direction

Adini

Moses

Ullman

1997

IEEE Trans. Pattern Anal. Machine Intell.

900

411

View full text Add to dashboard Cite

Abstract-A face recognition system must recognize a face from a novel image despite the variations between images of the same face. A common approach to overcoming image variations because of changes in the illumination conditions is to use image representations that are relatively insensitive to these variations. Examples of such representations are edge maps, image intensity derivatives, and images convolved with 2D Gabor-like filters. Here we present an empirical study that evaluates the sensitivity of these representations to changes in illumination, as well as viewpoint and facial expression. Our findings indicated that none of the representations considered is sufficient by itself to overcome image variations because of a change in the direction of illumination. Similar results were obtained for changes due to viewpoint and expression. Image representations that emphasized the horizontal features were found to be less sensitive to changes in the direction of illumination. However, systems based only on such representations failed to recognize up to 20 percent of the faces in our database. Humans performed considerably better under the same conditions. We discuss possible reasons for this superioriority and alternative methods for overcoming illumination effects in recognition.

show abstract

Section: Introductionmentioning

confidence: 99%

Face recognition: the problem of compensating for changes in illumination direction

Adini

Moses

Ullman

1997

IEEE Trans. Pattern Anal. Machine Intell.

900

411

View full text Add to dashboard Cite

show abstract

“…Far from modeling the HUB, Govindaraju et al [7][8][9][10] has detected head-and-shoulder based on the geometry properties of face profile. Prior information was required and their system has a high error rate.…”

Section: Research Backgroundmentioning

confidence: 99%

“…Many existing methods which use geometric modeling, boosted classifiers and SVM have been introduced [7][8][9][10][11][12]. But real-time applications, such as SL recognition systems, require a fast, compatible and synchronized method for HUB parts detection and tracking.…”

Section: Modeling the Human Upper Bodymentioning

confidence: 99%

Modeling of Human Upper Body for Sign Language Recognition

Bilal

Akmeliawati

Shafie

et al. 2011

The 5th International Conference on Automation, Robotics and Applications

View full text Add to dashboard Cite

Abstract-Sign Language Recognition systems require not only the hand motion trajectory to be classified but also facial features, Human Upper Body (HUB) and hand position with respect to other HUB parts. Head, face, forehead, shoulders and chest are very crucial parts that can carry a lot of positioning information of hand gestures in gesture classification. In this paper as the main contribution, a fast and robust search algorithm for HUB parts based on head size has been introduced for real time implementations. Scaling the extracted parts during body orientation was attained using partial estimation of face size. Tracking the extracted parts for front and side view was achieved using CAMSHIFT [24]. The outcome of the system makes it applicable for real-time applications such as Sign Languages Recognition (SLR) systems.

show abstract

“…They scan the original pages and binarize the images using a global threshold. Some techniques have been developed to locate faces using edge information on black and white images [2].…”

Section: Introductionmentioning

confidence: 99%

Recognition of quantized still face images

Chellappa

2009

2009 IEEE 3rd International Conference on Biometrics: Theory, Applications, and Systems

View full text Add to dashboard Cite

Abstract-In applications such as document understanding, only binary face images may be available as inputs to a face recognition (FR) algorithm. In this paper, we investigate the effects of the number of grey levels on PCA, multiple exemplar discriminant analysis (MEDA) and the elastic bunch graph matching (EBGM) FR algorithms. The inputs to these FR algorithms are quantized images (binary images or images with small number of grey levels) modified by distance and Box-Cox transforms. The performances of PCA and MEDA algorithms are at 87.66% for images in FRGC version 1 experiment 1 after they are thresholded and transformed while the EBGM algorithm achieves only 37.5%. In many document understanding applications, it is also required to verify a degraded low-quality image against a high-quality image, both of which are from the same source. For this problem, the performances of PCA and MEDA are stable when the images were degraded by noise, downsampling or different thresholding parameters.

show abstract

Locating human faces in newspaper photographs

Cited by 48 publications

References 11 publications

Face recognition: the problem of compensating for changes in illumination direction

Face recognition: the problem of compensating for changes in illumination direction

Modeling of Human Upper Body for Sign Language Recognition

Recognition of quantized still face images

Contact Info

Product

Resources

About