Ester Hait-Fraenkel

Ph.D. student

Thesis title: Multimodal image analysis using non-linear spectral representations

I am a Ph.D. student in the Electrical Engineering Department in the Technion – Israel Institute of Technology, in the field of image processing and computer vision,  under the supervision of Prof. Guy Gilboa

After receiving my B.Sc. in Electrical Engineering from the Technion in 2014 (Cum Laude), I received my M.Sc. in Electrical Engineering from the Technion in 2016, under the supervision of Prof. Guy Gilboa.

My M.Sc. research was in the subject of Blind Facial Image Quality Enhancement Using Non-Rigid Semantic Patches. In this work, we solved the practical general blind problem of enhancing the quality of facial images, in the case of multiple simultaneous unknown degradations. We combined problem-specific prior data, along with the original concept of non-rigid semantic patches of key facial features.

My Ph.D. research is in the subject of Multimodal Image Analysis using Non-linear Spectral Representations. In this framework, we presented the concept of Spectral Total-Variation Local Scale Signatures for Image Manipulation and Fusion. Now, following the concept of spectral representations, we explore Revealing Stable and Unstable Modes of Generic Denoisers through Nonlinear Eigenvalue Analysis. We also explore eigenfunctions of neural nets in Nonlinear Power Method for Computing Eigenvectors of Proximal Operators and Neural Networks.

PUBLICATIONS:
  1. Leon Bungert, Ester Hait-Fraenkel, Nicolas Papadakis, Guy Gilboa. “Nonlinear Power Method for Computing Eigenvectors of Proximal Operators and Neural Networks”, arXiv preprint arXiv:2003.04595 (2020).

  2. Hait-Fraenkel, Ester, and Guy Gilboa. “Revealing Stable and Unstable Modes of Generic Denoisers through Nonlinear Eigenvalue Analysis”, arXiv preprint arXiv:1909.12775 (2019).‏

  3. Hait, Ester, and Guy Gilboa. “Spectral Total-Variation Local Scale Signatures for Image Manipulation and Fusion.” IEEE Transactions on Image Processing, 28.2 (2019): 880 – 895.

  4. Hait, Ester, and Guy Gilboa. “Blind Facial Image Quality Enhancement using Non-Rigid Semantic Patches.” IEEE Transactions on Image Processing 26.6 (2017): 2705-2720. 

       Please follow this link to get to the M.Sc. project page (2016), including the TIP paper and Matlab code.​

CONFERENCE TALKS:
 
CONTACT ME AT:
etyhait@campus.technion.ac.il

© 2020 Vision & Sensing Lab