Comparison of face verification results on the XM2VTFS database

Matas, Jiřı́; Hamouz, M.; Jonsson, Kenneth; Kittler, Josef; Li, Y.; Kotropoulos, Constantine; Tefas, Anastasios; Pitas, Ioannis; Tan, T.; Yan, Hong; Smeraldi, Fabrizio; Bigün, Josef; Capdevielle, N.; Gerstner, Wulfram; Ben-Yacoub, Souheil; Abeljaoued, Y.; Mayoraz, Eddy

doi:10.1109/icpr.2000.903052

Cited by 78 publications

(64 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Four different institutions submitted results on the database which were subsequently published in [13]. This paper presents the results of a second contest using the same dataset and protocol, that has been organised as part of AVBPA 2003.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Face Verification Competition on the XM2VTS Database

Messer

Kittler

Sadeghi

et al. 2003

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. In the year 2000 a competition was organised to collect face verification results on an identical, publicly available data set using a standard evaluation protocol. The database used was the Xm2vts database along with the Lausanne protocol [14]. Four different institutions submitted results on the database which were subsequently published in [13]. Three years later, a second contest using the same dataset and protocol was organised as part of AVBPA 2003. This time round seven seperate institutions submitted results to the competition. This paper presents the results of the competition and shows that verification results on this protocol have increased in performance by a factor of 3.

show abstract

Section: Introductionmentioning

confidence: 99%

“…The database used was the Xm2vts database along with the Lausanne protocol [14]. Four different institutions submitted results on the database which were subsequently published in [13]. Three years later, a second contest using the same dataset and protocol was organised as part of AVBPA 2003.…”

mentioning

confidence: 99%

Face Verification Competition on the XM2VTS Database

Messer

Kittler

Sadeghi

et al. 2003

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…The development set is used uniquely to train the fusion classifier parameters, including the threshold (bias) parameter, whereas the evaluation set is used uniquely to evaluate the generalisation performance. They are in accordance to the two originally defined Lausanne Protocols [18]. The 32 fusion experiments have 400 (client accesses) × 32 (data sets)= 12,800 client accesses and 111,800 (impostor accesses) × 32 (data sets) = 3,577,600 impostor accesses.…”

Section: Database and Evaluationmentioning

confidence: 99%

A Novel Approach to Combining Client-Dependent and Confidence Information in Multimodal Biometrics

Poh

Bengio

2005

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. The issues of fusion with client-dependent and confidence information have been well studied separately in biometric authentication. In this study, we propose to take advantage of both sources of information in a discriminative framework. Initially, each source of information is processed on a per expert basis (plus on a per client basis for the first information and on a per example basis for the second information). Then, both sources of information are combined using a secondlevel classifier, across different experts. Although the formulation of such two-step solution is not new, the novelty lies in the way the sources of prior knowledge are incorporated prior to fusion using the second-level classifier. Because these two sources of information are of very different nature, one often needs to devise special algorithms to combine both information sources. Our framework that we call "Prior Knowledge Incorporation" has the advantage of using the standard machine learning algorithms. Based on 10×32 = 320 intramodal and multimodal fusion experiments carried out on the publicly available XM2VTS score-level fusion benchmark database, it is found that the generalisation performance of combining both information sources improves over using either or none of them, thus achieving a new state-of-the-art performance on this database.

show abstract

“…Then, the objective function to be optimized for classification is simplified to be the denominator of the objective function in (3). Denote the transformed class centers from the transformation matrix obtained in Section III-B2 as (12) Then, the corresponding objective function is (13) Denoting and , the optimal affine transformation matrix can be obtained by setting the derivative of the objective function to zero (14) Therefore, the optimal is (15) When the matrix is not of full rank, the inverse matrix can be replaced with the pseudo-inverse matrix. Finally, the transformation matrix is reset to be the product of a columnly orthogonal matrix and a diagonal matrix by using singular value decomposition as in (5) and (6).…”

Section: ) Initializationmentioning

confidence: 99%

“…FOR FACE VERIFICATION A large number of algorithms have been proposed for the face verification task; aside from techniques based on hidden Markov models (HMMs), the Bayesian method, support vector machines (SVMs), and neural networks (NNs) [13], [14], subspace learning techniques such as linear discriminant analysis (LDA) are the most popular. Motivated by the benefit to face verification of balancing class specific thresholds, in this work we develop a new subspace learning algorithm that can extract effective features for the verification task and at the same time ensure the balance of the class specific thresholds.…”

Section: Threshold Balanced Affine Transformationmentioning

confidence: 99%

Face Verification With Balanced Thresholds

Yan

Tang

2007

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Abstract-The process of face verification is guided by a prelearned global threshold, which, however, is often inconsistent with class-specific optimal thresholds. It is, hence, beneficial to pursue a balance of the class-specific thresholds in the model-learning stage. In this paper, we present a new dimensionality reduction algorithm tailored to the verification task that ensures threshold balance. This is achieved by the following aspects. First, feasibility is guaranteed by employing an affine transformation matrix, instead of the conventional projection matrix, for dimensionality reduction, and, hence, we call the proposed algorithm threshold balanced transformation (TBT). Then, the affine transformation matrix, constrained as the product of an orthogonal matrix and a diagonal matrix, is optimized to improve the threshold balance and classification capability in an iterative manner. Unlike most algorithms for face verification which are directly transplanted from face identification literature, TBT is specifically designed for face verification and clarifies the intrinsic distinction between these two tasks. Experiments on three benchmark face databases demonstrate that TBT significantly outperforms the state-of-the-art subspace techniques for face verification.

show abstract

Comparison of face verification results on the XM2VTFS database

Cited by 78 publications

References 8 publications

Face Verification Competition on the XM2VTS Database

Face Verification Competition on the XM2VTS Database

A Novel Approach to Combining Client-Dependent and Confidence Information in Multimodal Biometrics

Face Verification With Balanced Thresholds

Contact Info

Product

Resources

About