Jan Dorazil

This web page is evolving. More information about the papers, such as links to preprins and code, will be added soon.

About me:

I am a postdoc researcher at the Institute of Telecommunications, Brno University of Technology. My research revolves around Bayesian inference and machine learning applied to computer vision and image processing problems.

List of my publications:

• M. Jonák, J. Dorazil, V. Myška, R. Burget, a M. Kotrlý, „Classification of mineral phases in unpolished samples using a transformer neural network", in Proc. ICUMT, Florence, Italy, lis. 2025, s. 170–175.
  preprint  source code  synthetic data 
• J. Dorazil, B. H. Fleury, and F. Hlawatsch, ‘Fast Bayesian estimation using location-type variational representations’, in Proc. EUSIPCO, Italy, Palermo, May 2025.
  paper  source code 
• M. Luňáková, J. Dorazil, M. Jonák, Š. Ježek, and R. Burget, ‘Pen Ink Library: An interactive database of writing instruments based on Vis-NIR reflection spectra and optical properties of inks’, Forensic Sci. Int., May 2025, doi: 10.1016/j.forsciint.2025.112508.
  preprint 

• M. Jonák, J. Dorazil, M. Kolařík, Š. Ježek, R. Burget, and M. Kotrlý, ‘Forensic comparison of soil particles using Gaussian mixture models and likelihood ratio test’, in Proc. ICUMT, Ghent, Belgium, Oct. 2023, pp. 188–192. doi: 10.1109/ICUMT61075.2023.10333101.
• J. Dorazil, B. H. Fleury, and F. Hlawatsch, ‘Bayesian methods for optical flow estimation using a variational approximation, with applications to ultrasound’, in Proc. ICASSP, Rhodes, Greece, Jun. 2023, pp. 1–5. doi: 10.1109/ICASSP49357.2023.10095694.
• J. Dorazil, R. Repp, T. Kropfreiter, R. Prüller, K. Říha, and F. Hlawatsch, ‘Tracking carotid artery wall motion using an unscented Kalman filter and data fusion’, IEEE Access, vol. 8, pp. 222506–222519, Dec. 2020, doi: 10.1109/ACCESS.2020.3041796.
• J. Dorazil, R. Repp, T. Kropfreiter, R. Prüller, K. Říha, and F. Hlawatsch, ‘Feature drift resilient tracking of the carotid artery wall using unscented Kalman filtering with data fusion’, in Proc. ICASSP, Barcelona, Spain, May 2020, pp. 1095–1099. doi: 10.1109/ICASSP40776.2020.9054703.
• S. Dubey, R. P. Tripathi, M. K. Dutta, J. Dorazil, and P. Kříž, ‘Early detection of heart valve disease employing multiclass classifier’, presented at the Proc. ICUMT, Dublin, Ireland, Oct. 2019, pp. 1–6. doi: 10.1109/ICUMT48472.2019.8970870.
• J. Dorazil, R. Repp, T. Kropfreiter, R. Prüller, K. Říha, and F. Hlawatsch, ‘A multitarget tracking method for estimating carotid artery wall motion from ultrasound sequences’, in Proc. EUSIPCO, A Coruña, Spain, Sep. 2019, pp. 1–5. doi: 10.23919/EUSIPCO.2019.8902772.
• S. Srivastava, A. Singh, A. Yadav, M. K. Dutta, K. Říha, and J. Dorazil, ‘Automatic extraction of micro-aneurysms and hemorrhages from digital fundus image’, in Proc. TSP, Budapest, Hungary, Jul. 2019, pp. 249–254. doi: 10.1109/TSP.2019.8768847.
• J. Dorazil, K. Říha, and M. K. Dutta, ‘Common carotid artery wall localization in B-mode ultrasound images for initialization of artery wall tracking methods’, in Proc. TSP, Budapest, Hungary, Jul. 2019, pp. 605–608. doi: 10.1109/TSP.2019.8769077.
• P. Mangipudi, P. Bannerjee, M. K. Dutta, K. Říha, and J. Dorazil, ‘A stochastic method for ultrasound despeckling in images via dynamic weight allocation’, in Proc. TSP, Athens, Greece, Jul. 2018, pp. 228–231. doi: 10.1109/TSP.2018.8441447.
• J. Dorazil, K. Říha, M. K. Dutta, and A. Singh, ‘Change detection in presence of registration error using LTP histograms’, in Proc. TSP, Athens, Greece, Jul. 2018, pp. 380–383. doi: 10.1109/TSP.2018.8441176.