Professional History

Head of Innovation

Cortexica Vision Systems

 

June 2016 - Present

  • Exploring novel applications of visual search using emerging technologies such as augmented reality, virtual reality, machine learning and mobile heterogeneous processing (CPU, GPU, DSP and VPUs)
  • Exploring deep-learning applications for Android and iOS, e.g. optimising and deploying models for object detection and classification to solve challenging problems that would previously be extremely difficult to solve by conventional machine vision approaches
  • Collaborating with research, engineering and sales departments and encouraging innovation culture
  • Presenting at conferences, running webinars, authoring blog articles, securing new sources of funding and establishing new networks and collaborations
  • Rapid prototyping in C++, C#, QML and Java often using OpenCV, Qt, Unreal Engine and Unity as well as deep- learning frameworks such as Tensorflow, Caffe, Torch, OpenVino, etc.
  • Developed the initial prototype for action recognition, which led to a prestigious project with Hammerson 
  • Established partnerships with Axis and AAEON
  • Secured a letter of support from the European Space Agency for the OverPaSS satellite project
  • Created several edge compute hardware/software solutions and demos for a wide range of devices such as NVIDIA Jetson TX2, AAEON Boxer AI, Intel NUC, UP Squared, Cisco IOX devices, etc.
  • Created the world's first edge compute device for Personal Protective Equipment (PPE) ingress and real-time monitoring, which was also the first system utilising Intel Myriad X VPU. This product was first presented during the IoT Solutions World Congress (IOTSWC 2018) in Barcelona at both Intel and AAEON booths.

 

Senior Parallel Computing Software Engineer

Cortexica Vision Systems

 

February 2014 - May 2016

  • Rewritten the entire image processing pipeline using highly-optimised C++, CUDA and OpenCL code

  • Implemented multiple high-performance algorithms making the most out of AWS GPU instances

  • Implemented deep-learning applications using Torch framework, Lua and C++

  • Implemented Java-based automated ingestion system utilising several AWS features such as SQS, S3, EC1 auto- scaling using docker containers, etc.

  • Implemented several CUDA and OpenCL applications for mobile GPUs such as NVIDIA K1 or Samsung Exinos 5

  • Implemented couple of VR applications for HTC Vive and Oculus using Unreal Engine 4

  • Made several public talks focused on promoting the company and presenting its latest technology

 

Associate lecturer of GPU programming

Plymouth University

 

January 2012 - January 2014

  • Established NVIDIA-sponsored GPU computing lab at Plymouth University

  • Taught SOFT339 module focused on CUDA-based parallel computing and distributed systems

  • Led several 2-3 day long CUDA programming courses at various UK universities on behalf of NVIDIA

 

Education

 

PhD in GPU computing for cognitive robotics

Plymouth University

 

October 2008 - June 2014

  • Researched complex action and language acquisition in humanoid robots

  • Pioneered the application of GPU computing to cognitive robotics

  • Implemented Aquila: Software Architecture for Cognitive Robotics

  • Implemented MarsRoverSim: Simulator of Artificial Evolution of Mars Rovers

  • Taught SOFT130 module on software engineering (C# programming, UI design, coding standards) Supervised Barry Bentley during the “Evolution in robotic islands” project of The European Space Agency Supervised Ross Hammond during his final year project focused on extending Aquila's features

  • Attended a month-long training on iCub humanoid robot assembly and the use of its software stack

  • Attended RobotCub Summer School focused on learning of YARP robotic platform, iCub simulator and developing C++ applications for controlling iCub’s actions

 

BSc Computing

Plymouth University

 

September 2005 - May 2008

  • Awarded first class (Hons) degree and best student award

  • Placed on Dean's list for 1st, 2nd and 3rd year's performance

 

Life in the Universe

The University of Manchester

 

August 2006 - January 2007

  • Distance learning course focused on radio astronomy and the possibility of the life in the universe

 

Technical and informational systems

High Engineering School SPSS

 

September 1999 - June 2003

  • A-level equivalent subjects: computing, electronics, engineering, accounting, economy, English and Slovak

 

Qualifications

 

Certified ScrumMaster

Learning about agile scrum methodology and how to use it properly to increase the chances of team’s success

General Teaching Associate

Certificate of professional development, course run by Educational Development and Learning Technologies

 

Awards

 

Revell Research Systems Prize

Top Final Year Student in BSc (Hons) Computing

Best Papers

Postgraduate Conference for Computing: Applications and Theory (PCCAT) 2010 and 2011

 

 

Selected dissemination talks

 

 

Artificial Intelligence on the Edge - Safety

 

AAEON Webinar, 30 October, 2018

My dream about AI

 

TEDx Bratislava, Slovakia, 5 July, 2014 - Received standing ovation from around 1000 attendees.
English and Italian subtitles available.

Biologically-inspired Active Vision for Object Recognition

 

GPU Technology Conference, San Jose, California, 19 March 2013

Aquila: An Open-Source GPU-Accelerated Toolkit for Cognitive and Neuro-Robotics Research

 

NVIDIA GTC Webinar, 20 June, 2012

GPU-accelerated Action Acquisition Through Multiple Time Scales Recurrent Neural Network

 

GPU Technology Conference, San Jose, California, 19 March 2013

CUDA-based approach to cognitive robotics

 

NVIDIA Corp., Santa Clara, California, 29 September, 2011

Integrating Action and Language in Humanoid Robots

 

Slovak University of Technology, Bratislava, Slovakia, 4 July, 2011

Introduction to humanoid robot iCub, YARP and simulator

 

Plymouth University, UK, 11 June, 2010

Introduction to parallel computing using CUDA

 

Plymouth University, UK, 10 May, 2010

 

Press and Media Coverage

 

 

Engineering and robotics alumni - Martin Peniak

 

Plymouth University, 1 October, 2017

I migranti? Accogliamoli con i robot

 

La Scienze, 1 December, 2017

Slovenský vedec sa podieľa na výskume umelej inteligencie

 

Trend, 13 October, 2015

How the Booming GPU Computing Market Helped Turn an Immigrant’s Life Around

 

NVIDIA, 29 December, 2014

Meet the Good Will Hunting of the robotics department

 

Connect Magazine, 1 December, 2012

iCub poslúcha Martina na slovo: Tohto robota učí hovoriť Slovák!

 

Novy Cas, 1 August, 2012

Google Lunar X Contest Produces a Tegra Based Moon Rover

 

Netbook News, 7 February, 2012

PhD research highlighted in the keynote speech by NVIDIA CEO Jen-Hsun Huang

 

The International Conference for High Performance Computing, Networking, Storage and Analysis (SC11), 15 November 2011, Seattle, Washington

Evolving Autonomous Mars Rovers

 

ERCIM news, January 2011

NVIDIA CUDA Spotlight: Developing Robots with CUDA

 

NVIDIA, 26 December, 2010

AI and NVIDIA Parallel Processing: A PhD Student’s Research

 

NVIDIA, 15 December, 2010

Appeared on Geek Files TV Show

 

Sky Channel, 17 March, 2010

 

Grants and Donations

 

 

ARIADNA Project

European Space Agency, Evolution in robotic islands: Optimising the design of autonomous robot controllers for navigation and exploration of unknown environments, €24,466

Tesla C2075 x 2

NVIDIA, donation to support building a GPU cluster at Plymouth University

Tegra 3

NVIDIA, donation to support the development of lunar rover and lander control systems

Kayla ARM-based development

NVIDIA, donation to support building mobile GPU applications

Autonomous robot exploration in unknown terrains

euCognition Action Network, NA097-5, support for research, €5000

 

Publications

 

 

J. Zhong, M. Peniak, J. Tani, T. Ogata, A. Cangelosi (2017). Sensorimotor Input as a Language Generalisation Tool: A Neurorobotics Model for Generation and Generalisation of Noun-Verb Combinations with Sensorimotor Inputs. Autonomous Robots - Springer

B Sengupta, E Vazquez, M Sasdelli, Y Qian, M Peniak, L Netherton, G Delfino (2017). Large-scale image analysis using docker sandboxing. arXiv preprint arXiv:1703.02898

Broz, F., Nehaniv, C. L., Belpaeme, T., Bisio, A., Dautenhahn, K., Fadiga, L., Ferrauto, T., Fischer, K., Förster, F., Gigliotta, O., Griffiths, S., Lehmann, H., Lohan, K.S., Lyon, C., Marocco, D., Massera, G., Metta, G., Mohan, V., Morse, A., Nolfi, S., Nori, F., Peniak, M., Pitsch, K., Rohlfing, K.J., Sagerer, G., Sato, Y., Saunders, J., Schillingmann, L., Sciutti, A., Tikhanoff, V., Wrede, B., Zeschel, A., Cangelosi, A. (2014). The ITALK Project: A Developmental Robotics Approach to the Study of Individual, Social, and Linguistic Learning. Topics in cognitive science, 6(3), 534-544.

M. Peniak and A. Cangelosi (2015). Multiple Time Scales Recurrent Neural Network for Complex Action Acquisition: Model Enhancement with GPU-CUDA Processing. In: V. Kvasnicka, J. Pospíchal, S. Kozák ,P. Návrat, P. Paroulek (Eds.), Umelá inteligencia a kognitívna veda IV. Slovenská technická univerzita v Bratislave.

M. Peniak, A. Morse and A. Cangelosi (2013). Aquila 2.0: Software Architecture for Cognitive Robotics. International Joint Conference on Development and Learning (ICDL) and Epigenetic Robotics (ICDL-EPIROB) 2013. Osaka, Japan, August 18-22, 2013.

M. Peniak, D. Marocco, J, Tani, Y. Yamashita, K. Fischer and A. Cangelosi (2011). Multiple Time Scales Recurrent Neural Network for Complex Action Acquisition. International Joint Conference on Development and Learning and Epigenetic Robotics (ICDL-EPIROB), Frankfurt, Germany.

M. Peniak, A. Morse, C. Larcombe, S. Ramirez-Contla and A. Cangelosi (2011). Aquila: An Open-Source GPU- Accelerated Toolkit for Cognitive and Neuro-Robotics Research. International Joint Conference on Neural Networks (IJCNN), San Jose, California.

M. Peniak, B. Bentley, D. Marocco, A. Cangelosi, C. Ampatzis, D. Izzo, F. Biscani. (2010). An evolutionary approach to designing autonomous planetary rovers. TAROS-2010: Towards Autonomous RObotics Systems Conference, Plymouth, UK.

M. Peniak, B. Bentley, D. Marocco, A. Cangelosi, C. Ampatzis, D. Izzo, F. Biscani (2010). An island-model framework for evolving neuro-controllers for planetary rover control. International Joint Conference on Neural Networks (IJCNN), Barcelona, Spain.

M. Peniak, D. Marocco, S. Ramirez-Contla and A. Cangelosi (2009). An active vision system for navigating unknown environments: An evolutionary robotics approach for space research. Proceedings of IJCAI-09 Workshop on Artificial Intelligence in Space, Pasadena, California.

M. Peniak, D. Marocco and A. Cangelosi (2009). Co-evolving controller and sensing abilities in a simulated Mars Rover explorer. Proceedings of IEEE Congress on Evolutionary Computation (CEC), Trondheim, Norway.

M. Peniak, A. Cangelosi and D. Marocco (2008). Autonomous robot Exploration of unknown terrain: A preliminary model of Mars Rover robot. Proceedings of 10th ESA Workshop on Advanced Space Technologies for Robotics and Automation (ASTRA), Noordwijk, Netherlands.

 

Interests

 

Bodybuilding

Astrophotography

Agressive Skating

 

References

Angelo Cangelosi

Professor of Machine Learning & Robotics

Tony Belpaeme

Professor of Cognitive Systems and Robotics

Giorgio Metta

Vice Scientific Director

"Knowing others is wisdom, knowing yourself is enlightenment"