Blog

Blog

Uncategorized

Seminar on Artificial Intelligence

Department of Biomedical Engineering and AI Bay/DIH4.AI invite to seminar on 26th November 2020.

Program (CET):

  • 17.30-18.05: High Performance, Green Machine Learning for Embedded Edge Systems, David Gamba and Alicja Kwaśniewska, (SiMa.ai, CA, USA)
  • 18.05-18.40: ML and Data Classification, John Saunders, (Titus and Boldon James, UK)  
  • 18.40-19.15: AI in Practice: the “smarter” future ahead, Milos Manic (Virginia Commonwealth University, USA)

The event will be organized using Zoom platform. The invitation with Zoom link data will be sent two days before the seminar to all registered participants. Registration: https://aibay.ai/seminarium-z-cyklem-wykladow-26-listopada-2020/

 

About the speakers:

David Gamba – VP of Business Development at SiMa.ai (USA) – company focused on delivering a low power high performance ML platform accelerating performance of computer vision workloads for embedded edge applications (http://sima.ai/).

David has over 20 years experience in the semiconductor industry and has been recognized by the industry for introducing Hard Floating Point into FPGA devices as well as the OpenCL programming language into the programmable logic industry, which were key to driving the industry’s first wide scale adoption of FPGA accelerators in data centers.

Alicja Kwaśniewska – Software Architect at SiMa.ai specializing in machine learning applications for the embedded edge. She is also a co-organizer of the International Summer School on Deep Learning (dl-lab.eu) and has 20+ publications in the field, which were frequently awarded with e.g. best paper and best young professional paper awards. In her PhD dissertation she proposed innovative architectures of deep neural networks aimed at improving accuracy of remote medical diagnostics.

 

John SaundersSales Engineer at HelpSystems specializing in Data Identification, Classification and Protection. HelpSystems Data Security Suite combines Titus, Boldon James, Clearswift and GoAnywhere data security platforms to achieve a unified approach to managing and protecting your data.

Milos Manic – Professor of Computer Science at Virginia Commonwealth University (USA), director of VCU Cybersecurity Center, https://cybersecurity.vcu.edu/, Fellow Commonwealth Cyber Initiative (AI Assurance), https://cyberinitiative.org/,  officer at IEEE Industrial Electronics Society, an author of 200 research papers and the leader of more than 30 research projects. Research areas include: artificial intelligence, cyber security, critical infrastructure protection, energy security, resilient intelligent control, and other.

Projects

Cooperation with Intel on AI applications in thermal image…

Our Department and Intel (San Diego, USA) cooperate on the AI applications in thermal image processing of faces. The analyzed aspects are focused on face recognition, super-resolution algorithms and vital-signs estimation.

More information about our research at Intel blog:

Exploring Deep Learning Face Recognition with Thermal Images

and other websites:

Intel researchers propose AI that recognizes faces from thermal images

 

Events

IEEE Engineering in Medicine and Biology Conference – 7…

The Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) was organized in Berlin (23-27.07.2019). The Conference is the biggest and the most interesting conference in the area of Biomedical Engineering. The team from the Department of Biomedical Engineering has presented 7 papers during the event:

Can we rely on the pulse transit time – pressure relationship – models comparison *

Artur Poliński Adam Bujnowski Tomasz Kocejko Jerzy Wtorek

Publication Year: 2019, Page(s): 754 – 757

Respiratory signal of bathing person – preliminary study*

Adam Bujnowski Kamil Osinski Piotr Przystup Jerzy Wtorek

Publication Year: 2019, Page(s): 1274 – 1277

Methodology for assessing end-user requirements in the Ella4Life project: elders’ perspectives about self-monitoring *

Mariusz Kaczmarek Magdalena Velciu Luiza Spiru Adam Bujnowski Aleksey Andrushevich Edith Birrer

Publication Year: 2019, Page(s): 1409 – 1412

Evaluating Accuracy of Respiratory Rate Estimation from Super Resolved Thermal Imagery

Alicja Kwasniewska Maciej Szankin Jacek Ruminski Mariusz Kaczmarek

Publication Year: 2019, Page(s): 2744 – 2747

Evaluation of Facial Pulse Signals using Deep Neural Net Models

Jacek Rumiński Alicja Kwaśniewska Maciej Szankin Tomasz Kocejko Magdalena Mazur-Milecka

Publication Year: 2019, Page(s): 3399 – 3403

A wearable system developed to monitor people suffering from vasovagal syncope*

Michal Pietrewicz Artur Polinski Tomasz Kocejko Jerzy Wtorek

Publication Year: 2019, Page(s): 3584 – 3587

Electrodes array for contactless ECG measurement of a bathing person – a sensitivity analysis*

Kamil Osinski Adam Bujnowski Piotr Przystup Jerzy Wtorek

Publication Year: 2019, Page(s): 6583 – 6586

 

Congratulations to all authors!!!

 

Events

2nd International Summer School on Deep Learning (Gdansk, 1-5…

The 2nd International Summer School on Deep Learning (Gdansk, 1-5 July, 2019) organized by our team has just finished. More than 120 people from more that 20 countries participated in this event. This year again we were very lucky to host extraordinary speakers from many countries. Please visit the main website of the event (www.dl-lab.eu) or Twitter (https://twitter.com/hashtag/dliss19) for more information.

 

Awards

Awards during IEEE/IES HSI 2019 Conference

During the 12th International Conference on Human System Interaction IEEE HSI 2019 (Richmond, USA, June 25-27) teams from our department received BEST PAPER awards. Papers had been evaluated by reviewers before the conference and additionally presentations were evaluated during the conference. The following works presented by our team were awarded:

  1. Influence of Thermal Imagery Resolution on Accuracy of Deep Learning based Face Recognition (Maciej Szankin, Alicja Kwasniewska, Jacek Ruminski)
  2. Estimation of electrode contact in capacitive ECG measurement (Adam Bujnowski, Mariusz Kaczmarek, Jerzy Wtorek, Kamil Osinski, Domnika Strupinska)

Congratulations!

 

 

Uncategorized

Keynote lecture and 7 papers at 12th IEEE/IES HSI…

A team from our department participated in the 12th International Conference on Human System Interaction IEEE HSI 2019 (Richmond, USA, June 25-27). Prof. Jacek Rumiński presented the keynote lecture titled: Augmented Perception. Prof. Mariusz Kaczmarek, Dr. Adam Bujnowski and Dr. Tomasz Kocejko participated as experts in the very interesting discussion panels:

  • Panel 2: The Future of Brain-Computer Interfaces”: Dean Krusienski(VCU), KATHRYN Holloway (VA hospital), Tomasz Kocejko (Gdańsk University of Technology), Douglas P. Murphy(VAMC), Young IL Seo(VAMC),
  • Panel 3: Rehabilitation Technology and Human System Interaction”: Cang Ye (VCU), Mariusz Kaczmarek (GUT), Adam Bujnowski (GUT), Mizanoor Rahman (UWF), Paul A. Wetzel (VCU).

As a team we presented 7 accepted papers on:

  1. Detection of the Oocyte Orientation for the ICSI Method (Magdalena Mazur-Milecka, Emilia Kaczmarczyk, Lukasz Wrobel, Patryk Przybylski, Marika Trudnowska, Aleksandra Podwojcik, Monika Jagiello, Krzysztof Lukaszuk, Jacek Ruminski)
  2. Estimation of electrode contact in capacitive ECG measurement (Adam Bujnowski, Mariusz Kaczmarek, Jerzy Wtorek, Kamil Osinski, Domnika Strupinska)
  3. Optimal ECG lead for deriving respiratory signal (Piotr Przystup, Artur Polinski, Adam Bujnowski, Tomasz Kocejko, Jerzy Wtorek)
  4. Sensors integration in the smart home environment – a proposal to solve the problem with user identification (Mariusz Kaczmarek, Adam Bujnowski, Kamil Osiński, Artur Poliński, Tomasz Kocejko)
  5. Influence of Thermal Imagery Resolution on Accuracy of Deep Learning based Face Recognition (Maciej Szankin, Alicja Kwasniewska, Jacek Ruminski)
  6. Can Eye-tracking provide useful information on the skills acquisition by the radiology residents (Tomasz Kocejko, Tomasz Gorycki, Artur Polinski, Adma Bujnowski, Mariusz Kaczmarek, Jacek Ruminski, Antoni Nowakowski, Jerzy Wtorek)
  7. The influence of image masks definition on segmentation results of histopathological images using convolutional neural network (Kamil Janczyk, Tomasz Neumann, Jacek Rumiński).

Events

Special Session on Human-System Interaction, IECON 2018

We organized a special session on Human-System Interaction in Smart Environments during the IEEE International Conference on Industrial Electronics (IECON 2018). During the IECON 2018 Conference we presented 4 papers:

1. Towards Contactless, Hand Gestures-Based Control of Devices
2. Road Condition Evaluation using Fusion of Multiple Deep Models on Always-on Vision Processor
3. Smart weighing scale with feet-sampled ECG
4. Smart heating system for home extending utilization of renewable energy sources

Additionally, prof. Ruminski was invited to participate in the discussion panel focused on Smart Buildings.

In general, it was a very interesting conference and the participants were very interested in our works.

More information at: http://www.iecon2018.org/

Projects

Thermal Camera Modules (TCM) project

Biomedical Engineering Department is proud to announce start of the Thermal Camera Modules (TCM) project. First stage contains a “TCM Spark” module, which includes PCBA with a FLIR Lepton camera, a STM32F0 processor and series of interfaces available on the connectors (e.g. USB, UART, CAN). The dedicated PC software that will be able to capture the thermal images, as well as configure the cameras will be released together with the module. Two more modules: “TCM Fire” and “TCM Inferno” with extended functionality will be released in next stages. “Fire” will be equipped with a STM32F7 microcontroller and a 2.4inch TFT display, while “TCM Inferno” will be based on an IMX RT1050 microprocessor, a 4 inch TFT display and CMOS camera. Below, the specification for each module is presented.

TCM SPARK
– FLIR LEPTON sensor,
– STM32F0 microcontroller,
– microUSB connector,
– UART, CAN, SWDIO on ZIF connector

TCM FIRE
– FLIR LEPTON sensor,
– STM32F7 microcontroller,
– 256Mb external RAM,
– 2.4 inch TFT display,
– microUSB connector,
– UART, CAN, SWDIO on ZIF connector,
– FreeRTOS support

TCM INFERNO
– FLIR LEPTON sensor,
– IMX RT microprocessor,
– 256Mb external RAM,
– 4 inch TFT display,
– additional CMOS camera,
-UART, CAN, ETHERNET interfaces.

Events

IEEE EMB Conferences, Jeju (2017) & Honolulu (2018)

The International Conference of the IEEE Engineering in Medicine and Biology Society is probably the biggest and the most important conference focused on biomedical engineering. Last year the conference was organised in Jeju (Korea). Five participants from our departmenbt took part in this great event presenting 4 papers:

1. Automatic analysis of the aggressive behavior of laboratory animals using thermal video processing,
2. The motion influence on respiration rate estimation from low-resolution thermal sequences during attention focusing tasks,
3. A body position influence on ECG derived respiration,
4. A novel sensor for measuring temperature profile during the thermoablation,

This year, the EMBC 2018 conference was organized in Honolulu. We also presented 4 papers:
1. A Meta-Analysis of Pulse Arrival Time Based Blood Pressure Estimation
2. Long Distance Vital Signs Monitoring with Person Identification for Smart Home Solutions
3. Pose-Invariant Face Detection by Replacing Deep Neurons with Capsules for Thermal Imagery in Telemedicine
4. The Analysis of Temperature Changes of the Saliva Traces Left on the Fur During Laboratory Rats Soial Contacts

During the event we met Dr. John Webster. He is the editor of the: Medical instrumentation: application and design, Fourth Edition New York, John Wiley & Sons, 2009. He has also developed 22 other books including the Encyclopedia of medical devices and instrumentation, Second Edition, New York: John Wiley & Sons, 2006. We have been using his books for many years.

OLYMPUS DIGITAL CAMERA

We have also met our friends from Silesian University of Technology.

Next year, the EMBC 2019 will be organized in Berlin. We are looking forward to submit papers for this important event.

More photos from the event:

OLYMPUS DIGITAL CAMERA

Uncategorized

International Summer School on Deep Learning – 5 days…

The International Summer School on Deep Learning was organized by our Department between 2nd and 6th of July, 2018. It was very interesting event with 9 invited lectures, 14 mini-courses and many other attractions. The very interesting course was focused on the inference training using Intel/Movidius Neuro Stick. Each participant has received a stick sponsored by Intel and by the local organizers.
There was a great interest between potential participants. We could accept only 60 participants, but we have received much more applications. The participants were representing about 20 countries. During the event, there were many attractions like football competitions, deep learning challenge, “pierogi party”, Gdańsk sightseeing, and many other. Next year we will organize the second edition of the school. More information is presented at: www.dl-lab.eu.

Slides from the opening ceremony:

Example of activities after courses:

More information at: www.dl-lab.eu

More photos: