Philippe Burlina

[home] [about] [research] [recent] [papers] [CV] [other]

About

My research work is in machine vision and machine learning problems / algorithms / systems that are impactful for AI robustness / trust / assurance, including: AI fairness, privacy, adversarial machine learning, domain adaptation, low shot, anomaly detection. I lead teams and design AI algorithms that are impactful for applications of AI in computer vision, autonomy, and healthcare.

Contact

philippe [(dot)] burlina ([hat]) gee m ail ([dot]) c o m

Current Position

Principal scientist / JHU/APL Intelligent Systems Center

Associate Research Professor / Johns Hopkins University Dept. of Computer Science

Faculty, JHU MCEH, Malone Center for Engineering in Healthcare

Faculty / Johns Hopkins University School of Medicine, Wilmer Eye Institute

Previous Positions

Director, software development / FileNET (now part of IBM)

V.P. Engineering / eGrail (purchased by FileNET)

Co-founder and tech lead / ImageCorp (now part of SAIC/Leidos)

Areas of Work/Research

Machine vision, machine learning, machine intelligence, deep learning.

Assured AI, trusted AI, AI fairness and privacy, domain adaptation, low-shot learning, anomaly detection, adversarial machine learning.

Generative models.

Deep learning applications in autonomy, medical diagnostics, biomedical image analysis.

Enterprise software development.

Education

Ph.D. and M.S., Electrical Engineering, University of Maryland, College Park. (Computer Vision)

Diplôme d’Ingénieur, Université de Technologie de Compiègne, France. (Computer Science) (University of Pennsylvania, Philadelphia, PA., Electrical Engineering)

Tools, Languages, Frameworks

Python, PyTorch, Keras, OpenCV, Python data science libraries, C++, Matlab, PHP, MySQL, …

Patents

patents

Papers

most recent papers from google scholar

Sample Papers

W Paul, IJ Wang, F Alajaji, P Burlina, Unsupervised Discovery, Control, and Disentanglement of Semantic Attributes With Applications to Anomaly Detection, Neural Computation 33 (3), 802-826

Bo Hui, Yuchen Yang, Haolin Yuan, Philippe Burlina, Neil Gong, and Yinzhi Cao, Practical Blind Membership Inference Attack via Differential Comparisons, in the Proceedings of Network & Distributed System Security Symposium (NDSS), 2021.

N Joshi, P Burlina, AI Fairness via Domain Adaptation, arXiv preprint arXiv:2104.01109

N Drenkow, P Burlina, N Fendley, O Odoemene, J Markowitz, Addressing Visual Search in Open and Closed Set Settings, IEEE/CVF Conference on Computer Vision and Pattern Rec, Workshop Geo., 2021.

W Paul, A Hadzic, N Joshi, F Alajaji, P Burlina, TARA: Training and Representation Alteration for AI Fairness and Domain Generalization, arXiv preprint arXiv:2012.06387

N Drenkow, N Fendley, P Burlina, Random Projections for Adversarial Attack Detection, arXiv preprint arXiv:2012.06405

P Burlina, W Paul, P Mathew, N Joshi, KD Pacheco, NM Bressler, Low-shot deep learning of diabetic retinopathy with potential applications to address artificial intelligence bias in retinal diagnostics and rare ophthalmic diseases, JAMA ophthalmology 138 (10), 1070-1077 10 2020

PM Burlina, W Paul, PA Mathew, NJ Joshi, AW Rebman, JN Aucott, AI Progress in Skin Lesion Analysis, arXiv preprint arXiv:2009.13323, and COmputers in Bio and Medicine, 2020

N Fendley, M Lennon, IJ Wang, P Burlina, N Drenkow, Jacks of All Trades, Masters of None: Addressing Distributional Shift and Obtrusiveness via Transparent Patch Attacks, European Conference on Computer Vision, 105-119 2020

H Bhatia, W Paul, F Alajaji, B Gharesifard, P Burlina, R'{e} nyi Generative Adversarial Networks, arXiv preprint arXiv:2006.02479, Neural Computation, to appear, 2021.