Conference Program
-
- Day 1 23 June 2025
- Day 2 24 June 2025
- Day 3 25 June 2025
- Day 4 26 June 2025
- Auditorium
- Breakout Rooms
- Foyer
Keynote: Empowering Autonomy at Home: DIY Cognitive Assistance in Smart Homes
Sylvain Giroux
Depth cameras for frailty assessment: a dataset for automatic balance tests evaluation
Mario Quinde, Bruno Rioja, Oscar Belmonte-Fernandez, Gerardo Castillo, Miguel Castro
On the Opportunities and Risks of Generative AI in Mental Health
Shweta Premanandan, Jennifer Gross, Sofia Ouhbi
Towards Mobile Deep Venous Function Assessment: An Algorithmic Approach
Vincent Abt, Sergio Staab, Florian Kirchbuchner, Arjan Kuijper
Towards a Unified Architecture for Remote Patient Monitoring
Federico Bergamini, Giovanni Donato Gallo, Daniela Micucci
Chairs: Juan Carlos Augusto, Anton Gradišek
Ensuring the robustness and providing explanations for Artificial Intelligence (AI) solutions applied in the industry is imperative, particularly in the context of securing trustworthy AI. Increasingly complex real-world applications of AI utilise black-box models based on deep learning approaches to demonstrate high predictive accuracy and enhance industrial process efficiency. However, the decisions made by these black-box models are often difficult for human experts to understand – and, consequently, to act upon. The complete action plan to be performed based on, for example, the detected symptoms of damage and wear often requires complex reasoning and planning processes, involving many actors and balancing different priorities. Thus, operators, technicians and managers require insights to understand what is happening, why it is happening, what is the uncertainty in the observation, and how to react. The effectiveness of an industrial system hinges on the relevance of the actions undertaken by operators in response to the alarms. Therefore, establishing trustworthy AI involves not only accurate detection but also the provision of understandable, reliable, and comprehensive insights to facilitate informed decision-making and enhance the overall performance and robustness of the industrial system, which can be demonstrated by the adaptiveness and efficient decision-making in complex and fast-changing environments.
On arrival at the registration desk, you will receive your conference materials and name badge. These will give you access to all scheduled sessions, workshops, and social events. We are delighted to have you with us and look forward to inspiring discussions, new insights, and valuable networking opportunities. If you have any questions, our staff at the registration desk will be happy to help.
A chance to pause, refresh, and connect with fellow participants. Use this time to exchange ideas and continue the conversation beyond the sessions. Coffee and light refreshments will be provided.
Explore the Fraunhofer IGD and discover practical applications of our research. In a series of short demonstrations, our experts will showcase innovative technologies and solutions developed at the institute.
Kick off your conference experience with a warm welcome! Make valuable contacts at an exclusive networking event at the Fraunhofer IGD, accompanied by exquisite drinks and canapés. The event is open to all full conference and workshop/tutorial participants.
- Auditorium
- Breakout Rooms
- Foyer
Keynote I: Privacy Preserving in Edge Intelligence: Evaluation and Insights
Bahman Javadi
Keynote II: Accurate and Cost-Scalable Visual-Geometric Matching for indoor IoT Services
Junji Takahashi
Rule-Based Adaptive Energy Optimization for Neural Network-Based Energy-Efficient Transmission in Intelligent Systems
David Emmanuel, Nimesha Gamage, Anil Fernando
Quality Model Tool for Intelligent Environments: Initial Findings from a Literature Review
Mrinal Sharma
CAWN: Calibration-less Wi-Fi Sensing with NeRF
Natsumi Yoda, Ryota Kokubu, Yui Yamashita, Yuusuke Kawakita, Yoshito Tobe
Middleware for Assisted Living Homes (MiddAALware)
Omer Faruk Kacar
Chairs: Adityarajsingh Santokhee, Juan Carlos Augusto, Carlos Rodriguez Dominguez
Keynote: From ambient to space and back. . . with a little help from machine learning
Bogdan Ruszszak
The Role of semantic Models in constraining Pattern Recognition in Modern AI Systems
Felix Schaller
Vertical Projected Area Measurement Method Using Semantic Segmentation for Crop Growth Prediction
Rintaro ITO, Nobutada Fujii, Tekehide Soh, Shinichi Nakano
Leveraging Artificial Intelligence to Forecast Bird Sightings and its Visualization with Augmented Web Technologies
Juan Morales-Garcia, Paula Gonzalez-Martínez, Jose Melgarejo-Hernandez, Paula García-Tapia-Mateo, Irene Garriogos, Jose-Norberto Mazon
Configuring the transformer-based electricity market predictions using ant colony optimization
Jakub Sahs, Bogdan Ruszszak
Reinforcement Learning in de novo Molecular Design: Comparative Review of REINVENT 4, MolDQN, and ORGANIC
Arjun Subramanian, Connor Taylor
Chairs: Andrés Muñoz Ortega, Raquel Martínez-España, Fernando Terroso, Andrés Bueno-Crespo
In machine learning, missing features can disrupt predictive models. Active Feature Acquisition (AFA) addresses this by identifying which missing data to acquire under budget constraints. This tutorial introduces AFA for static and streaming data, guiding participants through implementing stream-based AFA methods. By the end, participants will have hands-on experience developing their own AFA solutions. This tutorial is designed for attendees unfamiliar with the topic but will also provide valuable insights for researchers working with data streams.
The development of Intelligent Environments presents significant complexity due to the diverse and advanced technologies involved. These systems integrate a range of components, including Sensor Networks, Artificial Intelligence algorithms, and user-personalised Human-Computer Interfaces. Each of these technologies is intricate on its own, and their combination in a cohesive system adds further layers of complexity. Additionally, there is no standard accepted way to build them which will guarantee good outcomes.
The aim of this tutorial is to introduce an enhanced approach to the User-Centred Intelligent Environment Development Process, referred to as UCIEDP2. This refined process aims to improve the development lifecycle of Intelligent Environment systems by emphasising quality assurance at every stage. UCIEDP2 focuses on integrating best practices, structured methodologies, and rigorous evaluation criteria to address the challenges associated with developing these complex systems. By providing a more systematic framework, UCIEDP2 seeks to ensure that the resulting Intelligent Environments are not only technically sound but also ethically viable for their intended users.
On arrival at the registration desk, you will receive your conference materials and name badge. These will give you access to all scheduled sessions, workshops, and social events. We are delighted to have you with us and look forward to inspiring discussions, new insights, and valuable networking opportunities. If you have any questions, our staff at the registration desk will be happy to help.
A chance to pause, refresh, and connect with fellow participants. Use this time to exchange ideas and continue the conversation beyond the sessions. Coffee and light refreshments will be provided.
Take some time to relax and recharge with a buffet lunch offered in the foyer. Enjoy a variety of dishes while engaging in informal discussions with fellow attendees. It’s a perfect moment to reflect on the morning sessions and make new connections.
A chance to pause, refresh, and connect with fellow participants. Use this time to exchange ideas and continue the conversation beyond the sessions. Coffee and light refreshments will be provided.
- Auditorium
- Foyer
Join us as we officially begin the main conference of IE2025, following two days of dynamic workshops. This session will highlight the central themes of the conference and outline what lies ahead. Hear welcoming remarks from the organizers and gain insight into the key topics that will shape the coming days.
Edge computing is more and more used for running machine learning algorithms to assist IoT (Internet of Things) tasks such as video analytics, compression, anomaly detections, privacy preservation, and others. On the other hand, edge computing is often resource constrained, challenging training and inference phases of machine learning algorithms.
In this talk, we will discuss Federated Learning (FL), a machine learning paradigm that enables a cluster of decentralized edge devices to collaboratively train a shared machine learning model without exposing users’ raw data and hence providing privacy to clients. However, the intensive model training computation is energy-demanding and poses severe challenges to end devices’ battery life. We will discuss few approaches to develop a training pace controller deployed on the edge devices that actuates the hardware operational frequencies over multiple configurations to achieve energy-efficient federated learning; and to tackle the straggler problem in FL via the decentralized selection of coresets, representative subsets of a dataset, where our approach creates coresets directly on edges and optimizes the coreset clusters to reduce FL training time and hence energy and other resource usage.
We will conclude with discussing further challenges of machine learning on edge devices.
The Meliferea chatbot: Harnessing large language models to interact with geoanalytical services using natural human language
Andreas Kamilaris, Valentijn Embrecht
Integrated Underwater Data Transmission and Object Detection System Using TinyML and Multi-Hop Networks
Ch Madhu Bhushan, Shahedul Haque Laskar, Jahnava Sai Garikapati, Sai Srihitha Polapragada, S Hemanth Durga Kumar, Firoj Gazi, MD Muzakkir Hussain
Leveraging Edge Resources for Indoor Localization for Improved Accuracy
Meka Naga Nandini Devi, MD Muzakkir Hussain, Shahedul Haque Laskar, Firoj Gazi
Exploring Tradeoffs of Annotation Cost and Model Accuracy with Contrastive Learning for Yoga Pose Classification
Damiran, Zolboo; Matsui, Tomokazu; Matsuda, Yuki; Suwa, Hirohiko; Yasumoto, Keiichi
Human Activity Recognition in Smart Homes Using IMU Wristbands: A Comparative Study of Sampling Strategies and Machine Learning Models
Chabih, Nada; Rice, Keven; Bouchard, Kevin; Gaboury, Sébastien; Maitre, Julien
Improving Sleep Quality Classification Through Synthetic Sensor-Based Medical Data Generation
Massa, Silvia; Fenu, Matteo
In the past decade, there has been significant progress in private space development, commonly referred to as NewSpace. One of the most notable advancements is the deployment of satellite constellations, with numerous small satellites being placed in low Earth orbit (LEO) to support communication and remote sensing applications. Looking ahead, it is anticipated that CPS (Cyber-Physical Systems) will be extended into space by leveraging artificial satellites as IoT devices or NTN communication platforms. To achieve this vision, we are conducting fundamental research on high-bandwidth communication satellites utilizing formation flying techniques. This presentation will introduce these emerging trends and highlight the related research activities.
AI-Driven Differential Diagnosis: Leveraging RAG and LLMs in Intelligent Healthcare Systems
Pinna, Simone; Massa, Silvia Maria; Riboni, Daniele
Parkinson’s Disease Freezing of Gait (FoG) Symptom Detection Using Machine Learning from Wearable Sensor Data
Hasan, Mahmudul
On arrival at the registration desk, you will receive your conference materials and name badge. These will give you access to all scheduled sessions, workshops, and social events. We are delighted to have you with us and look forward to inspiring discussions, new insights, and valuable networking opportunities. If you have any questions, our staff at the registration desk will be happy to help.
A chance to pause, refresh, and connect with fellow participants. Use this time to exchange ideas and continue the conversation beyond the sessions. Coffee and light refreshments will be provided.
Take some time to relax and recharge with a buffet lunch offered in the foyer. Enjoy a variety of dishes while engaging in informal discussions with fellow attendees. It’s a perfect moment to reflect on the morning sessions and make new connections.
A chance to pause, refresh, and connect with fellow participants. Use this time to exchange ideas and continue the conversation beyond the sessions. Coffee and light refreshments will be provided.
Experience Darmstadt’s cultural gem with a guided tour of Mathildenhöhe.
In 1899, Grand Duke Ernst Ludwig founded the artists’ colony on Mathildenhöhe, helping to establish Darmstadt as a center for Art Nouveau and modern European art. The colony aimed to bring art and everyday life closer together. Its work reflects early Modernism through innovative architecture, spatial design, and artistic experimentation.
Mathildenhöhe Darmstadt was added to the UNESCO World Heritage List in July 2021. It includes the Wedding Tower, exhibition building, artists’ colony museum, plane tree grove, and residential houses – all significant examples of early 20th-century experimental architecture.
As part of the tour, we will take the elevator to the top of the Wedding Tower, where you can enjoy panoramic views over Darmstadt, Frankfurt, the Odenwald forest, and the Bergstraße region.
In the evening, we invite you to dinner at Restaurant Mathildenhöhe, located in the heart of the World Heritage Site. With views of the plane tree grove, the Wedding Tower, the Russian Chapel, and the Lily Pond, the setting offers a relaxed and elegant atmosphere. The evening provides a pleasant opportunity to network and connect with fellow guests.
Would you like to bring guests? Additional tickets are available for purchase so you can share the experience with friends and family.
- Auditorium
- Foyer
This morning’s opening will set the tone for a day filled with engaging talks and discussions. Let’s make the most of this concluding day together.
Large-scale distributed machine learning training networks are increasingly facing scaling problems with respect to FLOPS per deployed compute node. Communication bottlenecks can inhibit the effective utilization of expensive GPU resources. The root cause of these performance problems is not insufficient transmission speed or slow servers; it is the structure of the distributed computing and the communication characteristics it incurs. Large machine learning workloads typically provide relatively asymmetric, and sometimes centralized, communication structures, such as gradient aggregation and model update distribution. Even when training networks are less centralized, the amount of data that needs to be sent to aggregate several thousand input values through collective communication functions such as AllReduce can lead to Incast problems that overload network resources and servers. This talk discusses opportunities and challenges for a systems approach towards making distributed machine learner faster, more energy-efficient, and scalable.
Supporting Energy Consumption Prediction: A Sustainable Approach
Ziran, Zahra; Arman, Ala; Mecella, Massimo; Muzi, Francesco ; Piras, Giuseppe
Detecting Noise Pollution through a Spatio-temporal Deep Learning Model: A Case Study of Barcelona
Semper, Marc; Curado, Manuel; Vicent, Jose F.
Cost-effective ADAS for inter-vehicle distance estimation using computer vision and deep learning
Guerrero-Contreras, Gabriel; Balderas-Díaz, Sara; Díaz-Gomez, Alejandro; Medina-Bulo, Inmaculada; Domínguez-Jiménez, Juan José
Explainable Artificial Intelligence for Audio-based Detection of Emergency Vehicles
Balderas-Díaz, Sara; Guerrero-Contreras, Gabriel; Muñoz, Andrés; Durães, Dalila; Novais, Paulo
SPUR: Enhancing Cyclist Safety through Digital City Twins – A Data-Driven Infrastructure Planning and VR Simulation Approach
Michael Disser
Design of Immersive Environments to Enhance Learning for People with Disabilities
Camilla Poggianti
ACCESS: Accessibility to Clinical Care for People with ASD through Anxiety Management
Elisa Salatti
Forecasting Sterility Mosaic Disease in Pigeonpea Using Dynamic Bayesian Networks and 3D Point Cloud High-throughput Scanning Platform
Mikes, Vojtech; Kocian, Alexander; Kholova, Jana; Masner, Jan; Kleczkowski, Adam; Sharma, Mamta; Chessa, Stefano; Galba, Alexander; Simek, Pavel
Protecting Endangered Birds with Edge-AI: Real-Time Detection of Invasive Cats in Natural Parks
Acebrón, Floreal; Rosas, Erika; Cano, Juan Carlos; Manzoni, Pietro; Sebastian, Esther; CECILIA, José María
Effectiveness of bird species identification using Birdnet: Case study at the La Mata coastal lagoon
Seye, Ousman; Sebastián-González, Esther; Ortiz, David; Calafate, Carlos T.; CECILIA, José María
Evaluation of time-series models for evapotranspiration prediction in smart agriculture
González, Martín; Casino, Virginia; Calafate, Carlos T.; López-Espín, José Juan; CECILIA, José María
Improving Multi-Camera View Recommendation with Temporal and Camera Embedding
Lee, Kuan-Ying; Zhou, Qian; Nahrstedt, Klara
Collaborative Lightweight LLM Agents for Daily Activity Summarization on Edge Devices
Inohara, Kentaro; Inohara, Kentaro; Yamaguchi, Hirozumi; Amano, Tatsuya; Rizk, Hamada; Rizk, Hamada
An Intelligent System for Healthcare Service Logging in a Medium-Sized Language
Smerkol, Maj; Susič, Rok; Jarc, Tomo; Gültekin Várkonyi, Gizem; Gradišek, Anton
CNN-Swin Backbones in Radar Object Detection for Autonomous Vehicles using Raw ADC Signals
Banwait, Iqbal; Zeller, Niclas; Alirezaie, Javad
In this session, we celebrate outstanding contributions to the field presented at IE2025. Several awards will be given in recognition of exceptional research, innovative ideas, and impactful presentations. For more details on the award categories and selection process, please visit ie2025.fraunhofer.de/awards.
A chance to pause, refresh, and connect with fellow participants. Use this time to exchange ideas and continue the conversation beyond the sessions. Coffee and light refreshments will be provided.
Take some time to relax and recharge with a buffet lunch offered in the foyer. Enjoy a variety of dishes while engaging in informal discussions with fellow attendees. It’s a perfect moment to reflect on the morning sessions and make new connections.
A chance to pause, refresh, and connect with fellow participants. Use this time to exchange ideas and continue the conversation beyond the sessions. Coffee and light refreshments will be provided.
