Keynote Lectures

Abstract
In recent years, 3D human activity recognition and tracking has become an important topic in human-computer interaction. To preserve the privacy of users, there is considerable interest in techniques without using a video camera. In this talk, we first present RFID-Pose, a vision-assisted 3D human pose estimation system based on deep learning (DL). The performance of DL models depends on the availability of sufficient high-quality radio frequency (RF) data, which is more difficult and expensive to collect than other types of data. To overcome this obstacle, in the second part of this talk, we present generative AI approaches to generate labeled synthetic RF data for multiple wireless sensing platforms, such as WiFi, RFID, and mmWave radar, including a conditional Recurrent Generative Adversarial Network (R-GAN) approach and diffusion/latent diffusion based approaches. Next, we propose a novel framework that leverages latent diffusion transformers to synthesize high quality RF data, as well as a latent diffusion transformer with cross-attention conditioning to accurately infer missing joints in skeletal poses, completing full 25-joint configurations from partial (i.e., 12-joint) inputs utilizing received RF sensory data. Finally, we present our recent work TF-Diff, a novel training-free diffusion framework for cross-domain radio frequency (RF)-based human activity recognition (HAR) system, which enables effective adaptation with minimal target-domain data.

Abstract
This keynote explores the transformative potential of the Internet of Things (IoT) in shaping the future of intelligent environments. It begins by introducing the foundational principles of IoT and its diverse applications across key verticals such as smart cities, healthcare, industrial automation, and precision agriculture. Through real-world projects, the session highlights innovative IoT-driven solutions that address complex challenges, demonstrating how IoT is revolutionizing industries and improving environmental sustainability and quality of life on a global scale. The discussion then transitions to the unprecedented opportunities created by the convergence of IoT with cutting-edge technologies. Artificial Intelligence (AI) is enabling IoT systems to evolve into autonomous, predictive, and adaptive ecosystems. The advent of 6G networks promises ultra-reliable, low-latency communication and massive connectivity, unlocking new possibilities for large-scale IoT deployments. Meanwhile, quantum technologies are poised to redefine the boundaries of security, computational power, and data processing in IoT frameworks. This keynote concludes by addressing the critical challenges that lie ahead, including scalability, security, ethical considerations, and the need for sustainable innovation. By illuminating these frontiers and suggesting future research directions, this session inspires new ideas and collaborations to drive the next wave of IoT-powered intelligent environments.

Abstract
Telecommunications are no longer merely “postmen of bits.” However, when envisioning the extreme capabilities required for future networks such as 6G, we still tend to focus on metrics like latency, ultra-high throughput, or mobility. The real challenge, instead, lies in demonstrating that new network architectures can effectively bridge the gap between actual societal needs and available technologies. The most successful “services” will be those that provide concrete solutions to real-world problems integrating communications, storage, processing and even sensing. Within this perspective, and as part of the large national investment program Telecommunications for the Future (Restart Program) in Italy, the University of Pisa participated to the WATCHEDGE focused project that has implemented a high-TRL field trial in the area of wildlife monitoring. An intelligent sensor network has been deployed to observe the behavior of wild animals, embodying the cloud continuum at the foundation of the NET4AI paradigm - the transformation of telecommunications to support artificial intelligence. The system is designed to automatically identify species and track their movements in real time, without human intervention, enabling continuous monitoring of wildlife such as wild boars, fallow deer, and wolves. Operating in a hostile environment with extreme requirements, the system integrates AI-accelerated cameras, edge computing technologies, energy harvesting through solar panels, and a hybrid communication infrastructure combining private 5G and satellite links. The field experiments made it possible to successfully test some initial implementations of semantic communications, aimed not only at compressing transmitted information - thus enabling the use of very low bit-rate networks - but also at improving reliability aspects, which are particularly critical in this domain. This project represents a significant step forward in applying artificial intelligence and the Internet of Things to biodiversity protection, contributing to environmental sustainability - a key element in the human-centric vision of Society 5.0. The data collected, both at the management and image levels, will constitute a valuable outcome of this on-field infrastructure, supporting advanced analytics and the long-term conservation of the natural environment.