| List |
| Topic: | Verifikace tváří s odhadem míry jistoty |
|---|---|
| Department: | Strojové učení |
| Supervisor: | Ing. Vojtěch Franc, Ph.D. |
| Announce as: | Bakalářská práce, Semestrální projekt |
| Description: | Performance of a face recognition (FR) system is influenced by a large set of factors characterizing input face images. The goal of the project will be to extend a given pre-trained FR system by a predictor estimating the performance of the FR system based on input face images. The performance predictor will be learned from mistakes the FR system makes in a test run. The learned performance predictor will be used for two purposes: i) to extend the FR system by the option to refrain from prediction in case the input faces have low-quality and ii) to compute optimal representation of a set of face images. Performance of the developed method will be quantitatively evaluated on face recognition tasks like face-verification and face-search using standard IJB benchmarks. |
| Bibliography: | - Klare at al. Pushing the Frontiers of Unconstrained Face Detection and Recognition: {IARPA} Janus Benchmark A. In proc. of CVPR. 2015.
- Best-Rowden et al. Learning Face Image Quality from Human Assessments. IEEE Trans. on Information Forensics and Security. 2018. - Abaza et al. Design and Evaluation of Photometric Image Quality Measures for Effective Face Recognition. IET Biometrics. 2014. |