If you are interested in learning how to create intelligent environments that can interact with users and enrich their activities, then you should join the IE2024 tutorials on intelligent environments. These tutorials will provide you with an overview of the main concepts, challenges, and opportunities in this multidisciplinary field. You will learn about state-of-the-art research and applications in areas such as sensors and actuators, the Internet of Things, signal processing (incl. audio and images), context awareness, artificial intelligence, human-computer interaction, software engineering, and pervasive and ubiquitous computing. You will also have the chance to interact with experts and peers and to get hands-on experience with some of the tools and platforms for developing intelligent environments.
Empowering Intelligent Environments: Sensing, Recognition, and Tracking
Sensing technology forms the backbone of IoT, and this tutorial provides an insightful overview of emerging sensing technologies in the realm of intelligent environments, focusing on Radio Frequency, LiDAR/depth cameras, and drones. We begin by briefly outlining the target applications and the advantages these sensing technologies bring. The tutorial then shifts focus to the technical challenges inherent in deriving meaningful insights from sensor data, such as contexts, locations, and building structures. Lastly, we delve into the role of AI in addressing these challenges, demonstrating how AI can be a powerful tool in transforming sensor data into actionable knowledge.
- Dr. Akira Uchiyama
- Dr. Akihito Hiromori
- Dr. Tatsuya Amano
- Dr. Hamada Rizk
Sensor-based Activity Recognition in Emerging Domains
Activity recognition using sensors has been a hot topic in ubiquitous computing. Advances in deep learning and sensor technology have expanded its application to fields like industrial and natural sciences. However, challenges arise due to insufficient training data and the complexity of activities. This tutorial introduces (1) state-of-the-art machine learning for practical activity recognition, (2) emerging domains like industrial settings and their datasets, and (3) its application in understanding animal behaviour in natural sciences.
- Dr. Takuya Maekawa
- Dr. Ryoma Otsuka
- Dr. Kei Tanigaki
- Dr. Qingxin Xia
- Dr. Naoya Yoshimura
Blockchain Applications and Development with an Intelligent Environment: An Easy Start
The development of smart contracts has extended the usage of blockchain technology from a simple ledger database for a decentralized cash transaction system into a platform that provides transaction intelligence, and further into a multifunctional state-sharing running platform for designing dependable decentralized applications. However, most current fast-booming blockchain platforms keep bringing new fashion features, which will bring the cost of breaking changes between different releases of the platform. This can be a barrier for developers, especially beginners, to start their development. In this tutorial, we aim to introduce the background of blockchain including consensus algorithms and smart contracts, and lay out how to develop a blockchain application with an Intelligent Development Environment through a set of hands-on online exercises.
- Dr. Weizhi Meng
- Dr. Wei-Yang Chiu
Privacy-aware learning for sensor-based affect and behaviour modelling
In an age characterized by data-driven frameworks, safeguarding privacy has become paramount, particularly in the context of Affective Computing and Human Activity Recognition, which enable a deeper understanding of human behavior and psychological states (e.g., emotions and cognitive load). Privacy is not merely theoretical but has practical implications, especially in domains like Affective Computing and Human Activity Recognition, where it is critical for protecting sensitive individual data and ensuring the security of health information or the anonymity of individuals in public surveillance. This tutorial should equip researchers and experienced practitioners with the necessary knowledge and methodological toolkit to adeptly navigate the complex landscape of Privacy-aware Learning. We will focus on recent techniques like Federated Learning, Split Learning, and Distributed Learning, offering hands-on experience in Privacy-Aware Learning. The tutorial also covers system-level optimization strategies to enhance the robustness and efficiency of Privacy-Aware Learning, addressing real-world challenges such as communication overhead, energy efficiency, and handling noisy datasets.
- Dr. Martin Gjoreski
- Dr. Dario Fenoglio
- Dr. Hristijan Gjoreski
- Dr. Valentin Rakovic
- Dr. Daniel Denkovski
- Dr. Marija Poposka
- Dr. Zoran Hadzi-Velkov
Digital Agriculture: informatics, technologies and user-centered design
In these recent decades, technological advancements in agriculture have reached dizzying speeds, and there is the urgency in ICT4Agriculture to let the farmers and the new artifacts to co-evolve together. Information technologies such as IoT, cloud computing, and AI, in conjunction with disciplines related to cognitive science and Human-Computer Interaction can face this problem. The tutorial reviews both technological and user-centered perspectives, discusses tailoring key technologies to agriculture, emphasizing interoperability, and showcasing examples from recent studies. It also presents qualitative study based on experts interviews in this field, in order to foster discussion and sense making on this particular issue about innovation in agriculture.
- Prof. Stefano Chessa
- Dr. Alexander Kocian
- Dr. Silvia Torsi
TinyML for Sustainable Development
TinyML enables machine learning on low-power microcontrollers, democratizing access to performant AI in remote, resource-constrained settings. This revolutionary technology unlocks new possibilities for sustainable development and scientific research by increasing equitable access to on-device intelligence worldwide. The emergence of TinyML has opened up new possibilities for building smart, ultra low-power devices ideal for resource-constrained settings. In recent years, TinyML has attracted significant interest from researchers, developers, and industries for its potential to enable innovative applications in healthcare, agriculture, transportation, conservation, smart homes, and more. Though currently restricted in reach, TinyML intersects topics across computer science and engineering curricula, making it an impactful educational tool. This hands-on workshop focuses on TinyML applications, providing training on commercially available hardware optimized for embedded ML deployment.
- Dr. Marco Zennaro
- Dr. Rytis Paškauskas
- Dr. John Shawe-Taylor
Generating Highly Realistic Doppelgänger Avatars: Technical Methodologies and Psychological Considerations in Virtual Environments
In this 90-minute tutorial, participants will gain a technical understanding of the generation of realistic avatars and their usage in multidisciplinary research studies or projects involving virtual environments (VR/AR). This session will also provide an overview of the psychological principles essential for the development of realistic avatars and their utilisation in studying human behaviour for addressing biases, and more. Technical hands-on training will equip participants with the skills to generate realistic avatars using Character Creator 3 and the Headshot plugin to create a doppelgänger avatar as well, followed by a quick integration into Unity for animation, movement, and lip-syncing. Newer applications of AI such as voice cloning techniques will also be discussed. The scope of such avatars is vast, ranging from research applications to content creation for education as well as other commercial activities.
- Dr. Sameer Kishore
- Dr. Nishtha Lamba
To register for the tutorials, please visit the website and fill out the registration form (coming soon).
We look forward to seeing you at the IE2024 tutorial on intelligent environments. If you have any questions or comments, please contact us.