Introduction to ACHIEVE

Size: px

Start display at page:

Download "Introduction to ACHIEVE"

Lorena McCormick
5 years ago
Views:

1 1 Introduction to ACHIEVE An European Training Network based on the experience of EUNEVIS Ricardo Carmona Galán WASC 2017, Córdoba (Spain)

2 Summary 2 ACHIEVE: AdvanCed Hardware/Software Components for Integrated/ Embedded Vision SystEms Marie Skłodowska-Curie action (H2020 work programme) European Training Network: coordinated research and training programme 9 Early-stage researchers will follow individual research projects and programmed secondments Network-wide training in specific and transferable skills Exposure to industrial procedures and needs: oriented to transference of results and employment

3 Beneficiaries & Partner Organisations Beneficiary Country Dept./ Division /Laboratory Contact person Consejo Superior de Investigaciones Científicas Université Clermont Auvergne SPAIN Instituto de Microelectrónica de Sevilla (IMSE-CNM) Ricardo Carmona Galán FRANCE Institute Pascal François Berry Université Bourgogne Franche-Comté FRANCE Laboratoire Electronique, Informatique et Image Dominique Ginhac Università degli Studi di Udine ITALY Artificial Vision and Real-Time Systems Lab Christian Micheloni Universidade de Coimbra PORTUGAL Instituto de Sistemas e Robótica Helder Araújo Universiteit Gent BELGIUM Image Processing and Interpretation Research Group Wilfried Philips Imasenic SPAIN Renato Turchetta Partner Organisation Country Dept./ Division /Laboratory Contact person FLIR BELGIUM Wouter Favoreel Prefixa USA Miguel Arias-Estrada Kovilta FINLAND Jonne Poikonen Nvidia Ltd FRANCE Guillaume Barat 3

4 Proposal 4 Excellence Impact Implementation

5 Main scientific objective 5 Development of a distributed vision platform composed of networked, smart and efficient embedded vision systems It will be the basic infrastructure in application scenarios that demand cooperative vision based on the in-node processing and extraction of relevant information This de-centralized scheme renders the system scalable, easily deployable and resilient to partial failure.

6 Applications 6 intelligent environments in the public and private spaces seamless and privacy-aware monitoring of human activities automated security and safety systems remote and interventional healthcare interactive marketing navigation aids for robotic driver-less vehicles intelligent transportation systems.

7 Specific scientific objectives 7 1. Development of integrated sensing and processing chips that combine image capture with on-chip acceleration feature extraction/learning with a limited number of resources and under a restricted power budget. 2. Conception of a chip architecture that extracts 2D and 3D information at sensor level for an enriched description of the scene that can be shared and combined between groups of camera nodes working in cooperation. 3. Design of hardware accelerators that will permit the implementation of heavy-duty feature and representation learning and deep learning inference. 4. Conception of compact and efficient reconfigurable embedded vision systems, where local processing of visual information is combined with agile transmission of metadata and a careful power management. 5. Development of the cooperative vision algorithms that will operate on an enriched representation of the scene that can be locally shared by a set of nodes, allowing them to react collectively. 6. Discarding the concept of a central hub where all the data crunching is performed in favour of a scalable distributed processing system in which visual information drives dynamic adaptation and feedback to enhance users experience. 7. Introduction of scalable, easily deployable, always-on, visual monitoring methods that will be the basis for a new class of products and services.

8 Work packages 8 WP# WP Title Type Leader Start End ESR s SO s WP1 Management and coordination Management CSIC M1 M48 ALL WP2 Recruitment Management IMAS M1 M6 ALL WP3 Smart image sensors andvision chips Research CSIC M6 M42 1, 2, 5, 9 1,2 WP4 Embedded vision systems Research UCA M6 M42 ALL 3,4 WP5 Cooperative vision and networked AI Research UGENT M6 M42 4, 5, 7, 8 5 WP6 Autonomous navigation and smart traffic Research UCOIM M6 M42 7, 8 5,6,7 WP7 Smart monitoring of human activities Research UNIUD M6 M42 5 5,6,7 WP8 Network training activities Training UBFC M6 M42 ALL WP9 WP10 Dissemination, IPR and exploitation Responsible Research and Innovation Dissemination UGENT M1 M48 ALL Management CSIC M1 M48 ALL WP11 Ethical clearance Ethics CSIC M1 M6 ALL

9 Work packages 9

10 Individual Research Projects 10 No. Host Title ESR1 ESR2 CSIC CSIC Computational imaging for early acceleration of deep learning inference schemes applied to video analytics Design of a light intensity/tof sensor in CMOS technology and a 2D/3D smart camera ESR3 UCA Towards Low Energy Smart Cameras ESR3 UBFC Design of parallel processing architectures for embedded deep learning techniques ESR5 UNIUD Distributed tracking by means of video pipeline reconfiguration ESR6 UCOIM Methods for autonomous navigation and localization in traffic environments ESR7 UGENT Cooperative tracking and visual analytics ESR8 UGENT Cooperative camera scene modelling using metadata ESR9 IMAS Development of smart, low power CMOS Image Sensors (CIS) for efficient 3D detection of scene

11 Training objectives The development of core research skills that will be trained at the host institutions through excellent research and local training activities related with the local doctorate programmes. 2. The development of additional research skills delivered through the network interactions, i. e. a common framework for multilevel/ multidisciplinary optimization. ESR s will learn to and propagate specifications and limitations across the system. 3. The development of transferable skills that will help to develop a highly-employable professional profile. They will be developed during network-wide training events and complemented by the local doctoral programmes. 4. The development of transferable skills that are directly related with the transference of research results and its translation into products and services after the accurate analysis of markets and society challenges and needs. 5. The development of transferable skills related with the implementation of responsible research and innovation (RRI). These skills will be also trained in network-wide training events.

12 Network-Wide Training Events 12 No. Main Training Events & Conferences Leader Month Location TE1 Design of PCDPs at ACHiEVE-ETN Kick-off Meeting CSIC M2 Brussels TE2 Scientific course on CMOS image sensors CSIC M9 Coimbra TE3 Seminar on Sociological impact of surveillance technologies UCOIM M9 Coimbra TE4 PhD Forum within the ICDSC 2018 UNIUD M TE5 Scientific course in Feature Learning in Embedded Systems UCA M21 Rennes TE6 Scientific course Cooperative Multi-Camera Computer Vision UGENT M TE7 Seminar on RRI at ACHIEVE-ETN Midterm Review Meeting CSIC M25 Brussels TE8 Seminar on IPR, Licensing and Transference UBFC M33 Dijon TE9 Special session on CAS for the Vision-Enabled IoT CSIC M34 Seville TE10 Special session & demos Cooperative Multi-Camera Systems UGENT M TE11 Joint Academic/Industrial Workshop in Silicon Valley PREF M42 Menlo Park TE12 Evaluation of the training at ACHIEVE-ETN Final Meeting CSIC M48 Seville

13 Network Organization 13

14 Network Organization 14

15 Interaction between participant organisations 15 Member WP Participation ESR Training Event Organisation Participation in Governing Boards CSIC 3,4,5,6,7 1,2,5,9 2,7,9 PC,SB, SC, S&TB, RRIB UCA 3,4,5,6,7 3,4,5,6 5 SB, SC, S&TB, DIEB UBFC 3,4,5,6,7 3,4 8 SB, S&TB, RTB UNIUD 5,6,7 5 4 SB, S&TB, DIEB UCOIM 4,6 6 3 SB, S&TB, RRIB UGENT 4,5,6,7 7,8 6,10 SB, SC, S&TB, RTB IMAS 3, 4 2,5,9 SB, RTB FLIR 3,4,5,6 7, 8 SB, DIEB PREF 3, 7 4,6 11 SB KOVI 3, 5, 5 1 SB NVID 3,4,5,6,7 3 SB, RRIB

16 Planned secondments 16 ESR No. Host Secondments 1 CSIC KOVI 2 CSIC IMAS 3 UCA NVID 4 UBFC UCA, PREF 5 UNIUD UCA, CSIC 6 UCOIM UCA, PREF 7 UGENT FLIR 8 UGENT FLIR 9 IMAS CSIC

17 Researcher Recruitment Training Chronogram Months Management Dissemination and Public engagement ESR 1 S S S S S S ESR 2 S S S S S S ESR 3 S S S S S ESR 4 S S S S S S ESR 5 S S S S S S S S S S S S ESR 6 S S S S S S ESR 7 S S S S S S S S S S ESR 8 S S S S S S S S S S ESR 9 S S S S S S Workshop Conference Seminar Course Meetings K E Dissemination Public Engagement

Low-power smart imagers for vision-enabled wireless sensor networks and a case study

Low-power smart imagers for vision-enabled wireless sensor networks and a case study J. Fernández-Berni, R. Carmona-Galán, Á. Rodríguez-Vázquez Institute of Microelectronics of Seville (IMSE-CNM), CSIC