Andrei BURSUC Address: 1, Allée des Galants Courts, 91033, Evry, France Email: [email protected] Phone: +33 6 77 50 87 20 EDUCATION “Doctorat ParisTech” PhD MINES ParisTech, Paris, France, October 2011 - December 2012  Specialization: Real time informatics, robotics and automation.  Thesis title: “Object based visual content indexing and retrieval”  Advisors: Prof. Françoise PRETEUX and Prof. Titus ZAHARIA PhD candidate Univesité Paris Descartes, Paris, France, October 2009 - September 2011 Engineer’s degree POLITEHNICA University of Bucharest, Faculty of Electronics, Telecommunications and Information Technology, Bucharest, Romania, October 2004 - September 2009  Specialization: Networks and Telecommunications Software  Ranked 12th out of 248 students PROFESSIONAL EXPERIENCE Post-doctoral researcher Institut Mines-Télécom;Télécom SudParis, ARTEMIS Department, Evry, France, January 2013 - present ALICE project (Assistance for Better Mobility and Improved Cognition of Elderly Blind and Visual Impaired)  Research on algorithms for real-time obstacle detection on mobile devices wearable by visually impaired persons.  Design techniques for the recognition of objects of interest and city landmarks in pictures or videos captured with the cameras of mobile devices (leveraging on Google Street View data). SmartCity+ project  Elaboration of a text-based search engine for city related data (documents belonging to public institutions, hospitals, public transport, crowd-sourced user generated data) and user images.  Participation in the MediaEval evaluation campaign within the Retrieving Diverse Social Images Task R&D engineer Télécom SudParis, ARTEMIS Department, Evry, France, October 2009 – December 2012 Ubimedia Joint Lab between Institut Télécom and Alcatel-Lucent Bell Labs  Designed and implemented a web-platform for video navigation and content-based search. The platform relies on a modular architecture with workload distributed over multiple servers communicating via HTTP requests/responses. The interface exploits the HTML5 specifications and the MPEG-7 standard for the hierarchical navigation inside the videos.

 Elaborated a content-based video search strategy for retrieving instances of the same object in different videos starting from minimal object information given by the user. The query object is defined by the user through the selection of a region of interest on a video frame and the search strategy is based on a technique that aggregates segmented regions/superpixels and color information with interest points and gradient intensity descriptors. Posed the problem of object identification as an energy optimization and proposed multiple solutions (variants of Greedy, Simulated Annealing, GraphCut).  Elaborated a hybrid image representation based on interest points and segmented regions and formulated the object recognition problem as a graph matching task, leveraging on pair-wise affinities.  Designed a multi-modal search technique enhancing the retrieval of object instances in video content starting from textual queries launched on the web (Flickr, Google Images). Meaningful images and descriptors are automatically filtered from the noisy data and used to generate new descriptors to be used as visual queries on the video repository.  Participated in the TRECVID evaluation campaigns within the Instance Search Task (retrieval of a query object in a database of 470K video clips) Bachelor thesis internship Institut Mines-Télécom;Télécom SudParis, ARTEMIS Department, Evry, France, March - September 2009  Designed and developed a multimedia web-based search engine compatible with MPEG-7 metadata. Vacation trainee Schlumberger, Drilling and Measurements Division, Hassi Messaoud, Algeria, July - August 2008  Analyzed and optimized the maintenance and repair chain in a workshop for large equipment (drill bits, turbines, drilling pipes, large sensors). SUMMARY OF PROFESSIONAL QUALIFICATIONS  Good knowledge of C, C++, Java, Python and basic knowledge of javascript, jQuery, php.  Development experience with openCV, simpleCV, scikit-learn, matlab, Octave on Windows and Linux.  Strong knowledge of data structures and algorithms.  Knowledge of Windows and Linux.  Deep interest in computer vision and machine learning.  Bilingual proficiency in English and Romanian; full professional proficiency in French.  Excellent communication skills with an interest in teaching and sharing. VARIOUS ACTIVITIES  Volunteer teacher for Romanian children in Paris aged 3 - 12, October 2010 - present  Member of the Administrative Council of SEFI (European Society for Engineering Education), October 2010 – September 2013  Volunteering activity at BEST (Board of European Students of Technology), 2005 - 2010 o Organized, coordinated and participated in over 40 events within 20 countries o Vice-president for Corporate Relations of Local BEST Group Bucharest, 2007 - 2008 o Project coordinator of JobShop 2007, job fair gathering 102 companies PERSONAL INTERESTS Online: Computer vision, image processing, search engines, machine learning, Big Data, open source. Offline: Volunteering, engineering education, reading, hiking, traveling.