The quotient image: class-based re-rendering and recognition with varying illuminations

Shashua, Amnon; Riklin-Raviv, Tammy

doi:10.1109/34.908964

Cited by 526 publications

(258 citation statements)

References 29 publications

Supporting

Mentioning

258

Contrasting

Order By: Relevance

“…The proposed frontal-view classification algorithm was trained using an SVM on 2D images generated from the 3D faces in the USF 3D database [7]. By rotating the 3D models and projecting them into the image plane, we can synthesize the 2D face images at different viewing angles.…”

Section: A Frontal-view Classification Using Dense Sift Descriptorsmentioning

confidence: 99%

Automated attendance system using markov random field algorithm

Maradwar¹,

Suryawanshi²,

Panpaliya³

et al. 2014

International Journal of Advanced Research in Computer and Comm

View full text Add to dashboard Cite

Abstract:Computer dependent communication is based on a very important factor, which is Authentication. Human face recognition is an important branch of biometric verification and has a wide range of application. In this paper we describe a method for Automated Attendance System which is based on face recognition by Markov Random Field algorithm. The system describes a method for marking the student's attendance using proposed algorithm and also provides additional features like maintaining the student database and required information about them. There are many advantages of this system over traditional method for attendance. It eliminates the overhead of manually marking the attendance and reduces the time and efforts. Markov Random Field algorithm is Pose-Invariant that does not require manually selection of facial landmarks or head pose estimation. In order to improve the performance of our pose normalization method in face recognition, we also present an algorithm for classifying whether a given face image is at a frontal or non-frontal pose. Experimental results on different datasets are presented to demonstrate the effectiveness of the proposed approach

show abstract

Section: A Frontal-view Classification Using Dense Sift Descriptorsmentioning

confidence: 99%

Automated attendance system using markov random field algorithm

Maradwar¹,

Suryawanshi²,

Panpaliya³

et al. 2014

International Journal of Advanced Research in Computer and Comm

View full text Add to dashboard Cite

show abstract

“…Human face image belongs to a special class. The facial feature position of different person only varies in a local area [14][15]. If we know the position and orientation of a face, we can use the spatial constraint among facial features to obtain the approximate position of each facial feature.…”

Section: Low-level Feature Based Face Tracking In Advancementioning

confidence: 99%

Bayesian Network Enhanced Prediction for Multiple Facial Feature Tracking

Huang¹,

Su²

2005

ELCVIA

View full text Add to dashboard Cite

It is challenging to track multiple facial features simultaneously in video while rich facial expressions are presented in a human face. To accurately predict the positions of multiple facial features' contours is important and difficult. This paper proposes a multi-cue prediction model based tracking algorithm. In the prediction model, CAMSHIFT is used to track the face in video in advance, and facial features' spatial constraint is utilized to roughly obtain the positions of facial features. Second order autoregressive process (ARP) based dynamic model is combined with graphical model (Bayesian network) based dynamic model. Incorporating ARP's quickness into graphical model's accurateness, we obtain the fusion of the prediction. Finally the prediction model and the measurement model are integrated into the framework of Kalman filter. The experimental results show that our algorithm can accurately track multiple facial features with varied facial expressions.

show abstract

“…In order to deal with the illumination variations on the faces, many methods have been exploited [1,7,11,13,16]. Recently, Basri et al [1] and Ramamoorthi et al [13] presented that the appearance of a convex Lambertian object can be well represented with a 9-D linear subspace.…”

Section: Introductionmentioning

confidence: 99%

Face Samples Re-lighting for Detection Based on the Harmonic Images

Chen

Qing

et al. 2004

Advances in Multimedia Information Processing - PCM 2004

View full text Add to dashboard Cite

Abstract. Different environment illumination has a great impact on face detection. In this paper, we present a solution by the face relighting based on the harmonic images. The basic idea is that there exist nine harmonic images which can be derived from a 3D model of a face, and by which we can estimate the illumination coefficient of any face samples. To detect faces under the certain lighting conditions, we relight those original face samples to get more new face samples under the various possible lighting conditions by an illumination ratio image and then add them to the training set. By train a classifier based on Support Vector Machine (SVM), the experimental results turn out that the relighting subspace is effective during the detection under the diverse lighting conditions. We also use the relighting database to train an AdaBoost-based face detector and test it on the MIT+CMU frontal face test set. The experimental results show that the data collection can be efficiently speeded up by the proposed methods.

show abstract

The quotient image: class-based re-rendering and recognition with varying illuminations

Cited by 526 publications

References 29 publications

Automated attendance system using markov random field algorithm

Automated attendance system using markov random field algorithm

Bayesian Network Enhanced Prediction for Multiple Facial Feature Tracking

Face Samples Re-lighting for Detection Based on the Harmonic Images

Contact Info

Product

Resources

About