Motion Features from Lip Movement for Person Authentication

Faraj, Maycel Isaac; Bigün, Josef

doi:10.1109/icpr.2006.814

Cited by 21 publications

(18 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A Bayesian classifier is used for classification, obtaining an average recognition rate of 90% at a false alarm rate of 5%. In other work, a new motion based feature extraction technique for speaker identification using orientation estimation in 2D manifolds is reported [15]. The motion is estimated by computing the components of the structure tensor from which normal flows are extracted.…”

Section: A Related Workmentioning

confidence: 99%

Using a Discrete Hidden Markov Model Kernel for lip-based biometric identification

Travieso

Zhang

Miller

et al. 2014

Image and Vision Computing

View full text Add to dashboard Cite

Abstract-In this paper, a novel and effective lip-based biometric identification approach with the Discrete Hidden Markov Model Kernel (DHMMK) is developed. Lips are described by shape features (both geometrical and sequential) on two different grid layouts: rectangular and polar. These features are then specifically modeled by a DHMMK, and learnt by a support vector machine classifier. Our experiments are carried out in a ten-fold cross validation fashion on three different datasets, GPDS-ULPGC Face dataset, PIE Face Dataset and RaFD Face Dataset. Results show that our approach has achieved an average classification accuracy of 99.8% 97.13%, and 98.10%, using only two training images per class, on these three datasets, respectively. Our comparative studies further show that the DHMMK achieved a 53% improvement against the baseline HMM approach. The comparative ROC curves also confirm the efficacy of the proposed lip contour based biometrics learned by DHMMK We also show that the performance of linear and RBF SVM is comparable under the frame work of DHMMK.

show abstract

Section: A Related Workmentioning

confidence: 99%

Using a Discrete Hidden Markov Model Kernel for lip-based biometric identification

Travieso

Zhang

Miller

et al. 2014

Image and Vision Computing

View full text Add to dashboard Cite

show abstract

“…Optical flow is the most common and easy to extract visual feature. In [10] dense optical flow is first calculated, then these dense velocity vectors are quantized by allowing only 3 directions (0 0 , 45 0 ,-45 0 ), and only 20 values resulting in a feature vector of 40 parameters. These quantization values were obtained by fuzzy c-means clustering.…”

Section: Dynamicmentioning

confidence: 99%

Comparative analysis of lip features for person identification

Saeed

2010

Proceedings of the 8th International Conference on Frontiers of Information Technology

View full text Add to dashboard Cite

Humans unconsciously use visual cues from lip motion not only in aural conversations but also to communicate subtle emotions. In the past decades lip detection and segmentation has been extensively studied in various fields from speech recognition and human computer interaction to biometrics, but literature comparing visual lip features is almost non-existent. In this paper we present a comparative evaluation of different visual features from lips. Geometric and appearance based features were extracted and their relevance to identifying people was studied by feature selection methods.

show abstract

“…Hybrid Methods use a combination of static and dynamic information [6], [18]- [22]. [16] DYNAMIC TI + AUDIO XM2VTS 295 EER 2 SANCHEZ [23] DYNAMIC TD + FACE XM2VTS 295 HTER 2.62 SANCHEZ [23] DYNAMIC TD + AUDIO XM2VTS 295 HTER 0.70 SANCHEZ [23] DYNAMIC TD + FACE + AUDIO XM2VTS 295 HTER 0.66 SANCHEZ [23] DYNAMIC TD + 2FACE + 2AUDIO XM2VTS 295 HTER 0.15 ABDULLA [21] HYBRID(SHAPE AND INTENSITY) CUSTOM 35 EER 18.0 CETINGUL [19] HYBRID(TEXTURE AND MOTION) MVGL-AVD 50 EER 3.6 CETINGUL [18] STATIC(TEXTURE)+DYNAMIC+AUDIO MVGL-AVD 50 EER 0.4 JOURLIN [5] STATIC(SHAPE) + AUDIO M2VTS 37 HTER 1.65 Table I provides an overview of the performance of various lip-biometric systems that perform speaker verification using only lip features. Table II provides an overview of the performance of various lip-biometric systems that perform speaker verification using lip features fused with other biometric traits such as audio and face.…”

Section: Relevant Workmentioning

confidence: 99%

Local Ordinal Contrast Pattern Histograms for Spatiotemporal, Lip-Based Speaker Authentication

Chan

Goswami

Kittler

et al. 2012

IEEE Trans.Inform.Forensic Secur.

View full text Add to dashboard Cite

Lip region deformation during speech contains biometric information and is termed visual speech. This biometric information can be interpreted as being genetic or behavioural depending on whether static or dynamic features are extracted. In this paper, we use a texture descriptor called Local Ordinal Contrast Pattern (LOCP) with a dynamic texture representation called Three Orthogonal Planes (TOP) to represent both the appearance and dynamics features observed in visual speech. This feature representation, when used in standard speaker verification engines, is shown to improve the performance of the lip-biometric trait compared to the state-of-the-art. The best baseline state-of-the-art performance was Half Total Error Rate (HTER) of 13.35% for the XM2VTS database. We obtained HTER of less than 1%. The resilience of the LOCP texture descriptor to random image noise is also investigated. Finally, the effect of the amount of video information on speaker verification performance suggests that with the proposed approach, speaker identity can be verified with a much shorter biometric trait record than the length normally required for voice-based biometrics. In summary, the performance obtained is remarkable and suggests that there is enough discriminative information in the mouthregion to enable its use as a primary biometric trait.

show abstract

Motion Features from Lip Movement for Person Authentication

Cited by 21 publications

References 3 publications

Using a Discrete Hidden Markov Model Kernel for lip-based biometric identification

Using a Discrete Hidden Markov Model Kernel for lip-based biometric identification

Comparative analysis of lip features for person identification

Local Ordinal Contrast Pattern Histograms for Spatiotemporal, Lip-Based Speaker Authentication

Contact Info

Product

Resources

About