Reconstructing Occluded Facial Components using Support Vector Data Description

지지 벡터 데이터 기술을 이용한 가려진 얼굴 요소 복원

  • Received : 2009.12.24
  • Accepted : 2010.01.26
  • Published : 2010.04.15

Abstract

Even though face recognition researches have been developed for a long ago, there is no practical face recognition system in real life. It is caused by several real situations where non-facial components such as glasses, scarf, and hair occlude facial components while facial images in a face database are well designed. This occlusion decreases recognition performance. Previous approaches in recent years have tried to solve non-facial components but have not resulted in enough performance. In this paper, we propose a method to handle this problem based on support vector data description, which trains the hyperball in feature space to find the minimum distance estimating the approximated face. In order to evaluate its performance and validate the effectiveness of the proposed method, we make several experiments and the results show that the proposed method has a considerable effectiveness.

얼굴 인식 분야는 오래전부터 꾸준히 연구되어 왔지만, 아직도 실용적인 얼굴 인식은 이루어지지 않고 있다. 이는 실제 얼굴 인식 시스템의 입력 영상의 경우, 실험실에서 획득된 얼굴 영상과는 달리 안경이나 스카프, 헤어스타일 등에 의해서 가려진 얼굴 영상인 경우에 인식 성능이 매우 저하되는 것에 기인한다. 이러한 비 얼굴 요소를 처리하기 위해, 최근 수년간 다양한 방식의 비 얼굴 요소처리 방법이 있었으나, 만족할만한 성능을 보이지 못했다. 본 논문에서는, 최근 관련 방법 중에서 특징 공간에서 최소거리의 볼을 찾아 근사값을 추정하는 방식인 SVDD를 이용하는 비 얼굴 요소 복원 방법을 제안하고, 실험을 통해 성능을 평가한다. 제안 방법의 실효성을 검증하기 위해, 비얼굴 요소 부분을 점진적으로 증가시켜 복원하는 실험 등 을 통해 실험한 결과, 제안 방법은 상당한 수준의 실효성을지니고 있음을 확인하였다.

Keywords

References

  1. B.-W. Hwang and S.-W. Lee, "Reconstruction of Partially Damaged Face Images Based on a Morphable Face Model," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.25, no.3, pp.365-372, 2003. https://doi.org/10.1109/TPAMI.2003.1182099
  2. C. Zhang, J. Wang, N. Zhao, and D. Zhang, "Reconstruction and analysis of multi-pose face images based on nonlinear dimensionality reduction," Pattern Recognition, vol.37, no.2, pp.325-336, 2004. https://doi.org/10.1016/j.patcog.2003.07.005
  3. J. Huang, V. Blanz, and B. Heisele, "Face Recognition Using Component-Based SVM Classification and Morphable Models," Pattern Recognition with Support Vector Machines, Lecture Notes in Computer Science, vol.2388, pp.334-341, 2002.
  4. L. Goldmann, A. Rama, T. Sikora, and F. Tarres, "On the Detection and Localization of Facial Occlusions and its Use within Different Scenarios," 10th IEEE Workshop on Multimedia signal Processing, vol.8, no.10, pp.592-597, 2008.
  5. S.-W. Lee, J. Park, and S.-W. Lee , "Synthesis of Face Exemplars using Support Vector Data Description," Proc. of the KIISE Fall Conference, vol.32, no.2, pp.835-837, 2005. (in Korean)
  6. J. Park, D. Kang, J. Kim, J. T. Kwok, and I. W. Tsang, "Pattern De-Noising Based on Support Vector Data Description," Proc. of Intl. Joint Conf. on Neural Networks, Montreal, Canada, pp.49-953, 2005.
  7. S.-W. Lee and S.-W. Lee, "SVDD-based Illumination Compensation for Face Recognition," Advances in Biometrics, Lecture Notes in Computer Science, vol.4642, pp.154-162, 2007.
  8. J. Park, D. Kang, J. T. Kwok, S.-W. Lee, B.-W. Hwang, and S.-W. Lee, "Facial Image Reconstruction by SVDD-Based Pattern De-noising," Advances in Biometrics, Lecture Notes in Computer Science, vol.3832, pp.129-135, 2005.