Research Day Report
Posted on September 7, 2023
The D²iCE (Data-Driven Computer Engineering) research group is a specialized hub that focuses on the convergence of academia and industry to address real-world challenges using AI and ML. With a vision to harness the transformative potential of these cutting-edge technologies, we specialize in the methodical application of cutting-edge techniques to address contemporary issues. We were thrilled to host the second edition of the D²iCE Research Day, which took place on the 7th of September, 2023 at the Lero Tierney Building on the University of Limerick campus. The event served as a collaborative platform for researchers, scholars, industry experts, and students from diverse backgrounds, all bound together by a shared interest in the realms of Artificial Intelligence and Machine Learning.
The event aimed to achieve several key objectives:
Knowledge Sharing: The D²iCE Research Day is a platform for researchers to share their latest findings, insights, and methodologies. It fostered a spirit of collaboration and cross-pollination of ideas.
Networking: With experts from academia and industry under one roof, the event facilitated networking opportunities. Attendees had the chance to build connections that could potentially lead to collaborations.
Industry Insights: As AI and ML continue to transform industries, our research day featured talks and panel discussions with industry leaders, shedding light on the practical applications of AI and the challenges they face, providing valuable real-world perspectives.
Highlights of the Day
The Research Day was a vibrant and dynamic event, filled with engaging activities and thought-provoking discussions:
Keynote Addresses: A special note of appreciation goes out to Conor Morris, a case manager in the field of ICT at the UL Technology Transfer Office, who delivered a talk on “Intellectual Property Management.” Conor’s insightful discourse shed light on the intricate world of patents, copyrights, and more, enriching the research groups understanding of this important aspect of intellectual property commercialisation.
Research Showcase: During the poster session, an array of projects and findings were displayed by researchers from the D²iCE group. This poster session served as a platform for in-depth discussions on a wide range of topics, including but not limited to:
Visual Question Answering Models for Autonomous Vehicles
Automotive Tracking and Detection
Hardware Acceleration in Autonomous Vehicles
Medical Imaging Advancements
Trajectory Prediction for Autonomous Vehicles These insights, presented through posters and interactive demonstrations, not only showcased the breadth of research but also exemplified the potential for advancements in these diverse areas of AI and ML.
Industry attendees: We were joined by representatives from Valeo, AMCS, Collins Aerospace, Jaguar Land Rover, Motional, Black Box, and others. Their active participation and invaluable insights greatly enriched our event, contributing to a wealth of knowledge and expertise that will undoubtedly shape the future of AI and ML.
Student Involvement: we had several insightful talks delivered by our PhD students and researchers. The diverse topics covered included:
“The Binspect Project” by Cathaoir Agnew: An exploration into the intriguing world of the Binspect Project, shedding light on its significance and implications.
“Revisiting Modality Imbalance in Multimode Pedestrian Detection” by Arindam Das: An in-depth examination of the challenges and solutions related to modality imbalance in multimode pedestrian detection, offering valuable insights into this critical area.
“Advancing Driver-Assistance Systems with Deep Learning: Object Detection in the Thermal Domain” by John Ringrose: A journey into the application of deep learning in enhancing driver-assistance systems, with a specific focus on object detection in the thermal domain.
“Designing an ETL and AWS” by Zahraa Abdulhuessein: A discussion on the intricacies of designing an Extract, Transform, Load (ETL) system and leveraging Amazon Web Services (AWS) for data processing and storage.
The topic of discussion revolved around “Near Field iToF LIDAR Depth Improvement from a Limited Number of Shots.” Mena Nagiub presentation shed light on the innovative methods and strategies used to enhance depth perception with limited data points in the realm of LIDAR technology.
Last but certainly not least, Darragh Glavin shared his expertise in the realm of the mental health domain, and how machine learning is being used in this topic.
Whye Kit Fong from Motional gave a keynote talk on the background and development of the famous nuScenes dataset. nuScenes is used extensively by D²iCE automotive researchers, so it was fascinating to get a talk from the creators.
Our heartfelt thanks go out to the organizing committee comprised of dedicated individuals including Daniel Jakab, John Ringrose, Sushil Sharma, and Aryan Singh. Their tireless efforts ensured the success of our event. Thank you to all the guys at Lero for hosting us!
Photos from the event
Everyone at the D²iCE Research Day
Attendees are fascinated by our talks
Arindam Das came all the way from Chennai, India, to deliver his talk to the group
Cathaoir Agnew describes the Binspect project
Darragh Glavin on the applications of machine learning in predicting the mental health of people
Colin Fitzpatrick wraps up the day by discussing the intersection of AI and the Circular Economy