20 Staniford Street, Boston MA 340 Putnam Avenue, Cambridge MA

Transcription

1 Curriculum Vitae Date Prepared: May 22, 2014 Name: Office Address: Home Address: Matteo Tomasi Work Phone: Work Work FAX: Place of Birth: 20 Staniford Street, Boston MA 340 Putnam Avenue, Cambridge MA San Gavino Monreale (Italy) Education 2004 BD Electronic Engineering Universitá di Cagliari 2006 MSc Electronic Engineering Universitá di Cagliari 2008 MSc Computer Engineering and Networks Universidad de Granada 2010 PhD Computer Engineering and Networks Universidad de Granada (Eduardo Ros and Javier Diaz) Postdoctoral Training May PostDoctoral Engineering Schepens Eye Research Institute (Eli Peli, Gang Luo) May Research Fellowship Ophthalmology Harvard Medical School Editorial Activities Ad Hoc Reviewer - Journal of Systems Architecture - Image and Vision Computing - IEEE Design and Test on Computers - Journal of NeuroComputing - IEEE Transactions on Parallel and distributed systems - International Symposium on Visual Computing - Journal of Real Time Image Processing Report of Funded and Unfunded Projects Funding Information Past DRIVSCO Learning to Emulate Perception-Action Cycles in a Driving School Scenario 1

2 European Commission (FP6-IST-FET, contract ) The main goal was to devise, test and implement a strategy of how to combine adaptive learning mechanisms with conventional control, starting with a fully operational human-machine interfaced control system and arriving at a strongly improved, largely autonomous system. The final system acts in a proactive way using different predictive mechanisms SENSOPAC - SENSOrimotor structuring of Perception and Action for emergent Cognition European Commission (FP6-IST-CSI) The project has combined machine learning techniques and knowledge about biological perception-action mechanisms. SENSOPAC has developed a cerebellum-based control system to solve complex haptic/touch problems using active sensing DINAMvision - Sistema de visión dinámica en tiempo real y su aplicación en robótica, vehículos y biomedicina Hybrid HMD Current Spanish Government (DPI ) The main goal was the implementation of a dynamic vision system for real-time processing and its application to robotic, biomedicine and driving scenarios. NIH/ ZY12890 The major goal of this study is the development and test of a hybrid vision enhancement glasses based on an optical see-through head mounted display (HMD) that can be used in many situations by visually impaired people Completion of enrollment Development of a Vision Assistive Device for Veterans with TBI-Associated Visual Dysfunctions DM DMRDP In this project, we are developing an intelligent collision alerting system that can determine collision risk in terms of time to collision and collision point, and then give TBI patients auditory alerting messages. The core image processing methods mimic a collision judgment mechanism in the human visual system that we found in a previous study with patients with tunnel vision Completion of enrollment Rehabilitation of Visual and Perceptual Dysfunction after Severe Traumatic Brain Injury DM DMRDP The overall aim is to conduct preliminary evaluations of new rehabilitation strategies and new functional assessment methods for homonymous hemianopia (HH) and spatial neglect (SN), disabling visual and cognitive perception 2

3 conditions that commonly occur as a result of severe traumatic brain injury (TBI) and stroke. Both HH and SN prevent detection of objects on the affected side, resulting in unsafe walking and driving. Using realistic tasks in virtual environments representative of everyday mobility challenges we are evaluating a novel optical device expansion prism (EP) glasses - combined with a new computerized perceptual-motor training regimen in helping people with HH and SN detect and avoid obstacles on the affected side Indoor Navigation FP7 European Funding PI Navigation outdoor is a well consolidated field with the use of GPS systems. Unfortunately the same accuracy and usability is missed in indoor environment where the GPS has no signal. In this project we propose to develop a mobile navigation system that helps visually impaired and potentially all normally sighted to find their way in large buildings. The final prototype will be delivered as a smart phone application and use the embedded camera and motion sensors to estimate the user position. Report of Regional, National and International Invited Teaching and Presentations International 2009 Low level vision on Chip, European Workshop on Advanced Predictive Sensor-Motor Control, Juodkrante (Lithuania) Report of Scholarship Peer reviewed publications in print or other media Journals: 1- Vanegas M, Tomasi M, Díaz J, Ros E. Multi-port abstraction layer for FPGA intensive memory exploitation applications. Journal of Systems Architecture 2010; 56(9): Tomasi M, Barranco F, Vanegas M, Diaz J, Ros E. Fine grain pipeline architecture for high performance phase-based optical flow computation. Journal of System Architecture 2010; 56(11): Tomasi M, Vanegas M, Barranco F, Díaz J, Ros E. High-performance optical flow architecture based on a multi-scale, multi-orientation phase-based model. IEEE Trans. on Circuits and Systems for Video Technology 2010; 20(12): Tomasi M., Vanegas M., Barranco F., Díaz J. and Ros E., Massive parallel hardware architecture for multi-scale image structure, stereo, and optical flow computation, IEEE Trans. on Circuits and Systems for Video Technology 2012; 22(2): Barranco F, Tomasi M, Díaz J, Vanegas M, Ros E. Parallel architecture for hierarchical optical flow estimation based on FPGA. IEEE Trans. on VLSI 2012; 20(6):

4 6- Pauwels K., Tomasi M., Díaz J., Ros E. and Van Hulle M.M., "A Comparison of FPGA and GPU for Real-Time Phase-based Optical Flow, Stereo, and Local Image Features", IEEE Trans. on Computers 2012; 61(7): Tomasi M., Vanegas M., Barranco F., Diaz J., and Ros E., Real-time architecture for a robust multiscale stereo engine on FPGA, IEEE Trans. on VLSI 2012; 20 (12): Barranco F., Tomasi M., Díaz J., Vanegas M., Ros E.: Pipelined architecture for real-time costoptimized extraction of visual primitives based on FPGAs. Digital Signal Processing 2013; 23(2): Barranco, F., Tomasi, M., Vanegas, M., Diaz, J., Granados, S., Ros, E. Hierarchical architecture for motion and depth estimations based on color cues. Journal of Real-Time Image Processing 2012; Article in Press. 10- Tomasi, M., Pundlik, S., Luo G., FPGA DSP co-processing for feature tracking in smart video sensors. Journal of Real-Time Image Processing 2014; Article In Press. Conferences: 1- Tomasi M, Díaz J, Ros E. Real time architectures for moving-objects tracking. ARC2007, Lecture Notes in Computer Science 2007; pp Tomasi M, Vanegas M, Barranco F, Díaz J, Ros E. Arquitectura multiescala de cálculo de flujo óptico basado en la fase. IX Jornadas de Computación Reconfigurable y Aplicaciones. JCRA2009, 2009, pp Barranco F, Tomasi M, Vanegas M, Granados SM, Díaz J. Entorno software para visualización y configuración de procesamiento de imágenes en tiempo real con plataformas reconfigurables. IX Jornadas de Computación Reconfigurable y Aplicaciones. JCRA2009, 2009, pp Tomasi M, Vanegas M, Barranco F, Díaz J, Ros E. A novel architecture for a massively parallel low level vision processing engine on chip. Industrial Electronics (ISIE), 2010 IEEE International Symposium on, Bari (Italy) (4-7 July 2010), pp Barranco F, Tomasi M, Vanegas M, Díaz J, Ros E. Arquitectura para extracción de primitivas visuales de bajo nivel en un mismo chip en tiempo real", III Congreso Español de Informática (CEDI 2010), X Jornadas de Computación Reconfigurable y Aplicaciones (JCRA2010), 7-10 Septiembre Valencia (Spain). Pp ISBN Rubio L, Vanegas M, Tomasi M, Diaz J, Ros E. On-chip ego-motion estimation based on optical flow. 7th International Symposium on Applied Reconfigurable Computing (ARC11), March Belfast (Ireland). 4

5 7- Barranco F, Tomasi M, Diaz J, Ros E. Hierarchical Optical Flow Estimation Architecture using Color Cues. 7th International Symposium on Applied Reconfigurable Computing (ARC11), March Belfast (Ireland). 8- Tomasi M, Churchill J, Wiegand JP, Houston K, Peli E, Bowers AR, Luo G. Peripheral Prisms Increase Blindside Eye And Head Scanning Movements During Outdoor Walking In Hemianopes: Preliminary Results. ARVO 2013 Annual Meeting, May 5-9, Seattle (USA). 9- Pundlik S, Tomasi M, Luo G. Collision Detection for Visually Impaired from a Body-Mounted Camera. Computer Vision and Pattern Recognition, IEEE Conference on. CVPR 2013, June Portland (USA). 10- Tomasi M, Bowers AR, Peli E, Luo G. Compensatory gaze scanning by patients with hemianopia during outdoor walking. ARVO annual meeting. May Orlando FL (USA). 11- Pundlik S, Tomasi M, Luo G. Evaluation of a portable collision warning device for visually impaired patients in an indoor obstacle course. ARVO annual meeting. May Orlando FL (USA). Thesis June 2010 Pyramidal architecture for stereo vision and motion estimation in real-time FPGA-based devices Advisors: Eduardo Ros and Javier Díaz Narrative Report I acquired my interest on digital electronic and computer vision during my studies in the University of Cagliari (Italy) especially at the practice in laboratory. This aspect has been improved with my Master Thesis which focused on video processing of moving objects in portable devices. After the M.Sc. degree in 2006, my research interests moved definitely on Hardware Description Languages (HDL) and hardware design, Real-time systems, Reconfigurable Systems for computer vision and image processing. I spent almost five years at the Department of Computer Architectures and Technology (University of Granada, Spain). This period started with the award of an Erasmus Grant in 2006 and ended with the achievement of my PhD degree in Here I started to appreciate the research investigating on realtime processing for stereo vision and motion estimation in Field Programmable Gate Array (FPGA) devices. During my stay in this department, I collaborated on several European projects supervised by Eduardo Ros and Javier Díaz (my thesis advisors). This experience has improved significantly my research skills and team abilities thanks to a fruitful international environment with project meetings and reviews. The heavy load of the projects and the publications has reduced the time for teaching. My contribution to education in Granada has been limited to collaborations during Information Engineering courses. 5

6 Currently, my research field is applied, with a postdoctoral fellowship at Schepens Eye Research Institute (Harvard Medical School), to low vision rehabilitation. This implies the development of a new evaluation method: not only with computer oriented tests but also with the recruitment of visually impaired people. The final goal of this research is a great motivation for me. My present position focuses on image magnification, edge detection, stereo vision and collision detection problems. The introduction of Digital Signal Processor (DSP) devices complements my real-time processing abilities with new hardware/software co-designs. During this time at SERI, I m also exploring the field of neuroscience including problems related with eye and head movements in walking scenarios. On summary, all acquired concepts involving computer vision in portable devices are now very important for the application in low vision aids. The current fellowship offers a potential link for future collaborations between US and Europe. Possible prototypes and patents can also represent the natural evolution of this work in the industrial world. 6