A new chapter begins! 🤩 I have recently joined the Interdisciplinary Transformation University of Austria (IT:U) in Linz as a Postdoctoral Researcher of Embedded ML and Smart Sensors, working with Prof. Michele Magno, within the Smart Space Sensing and Systems Lab (S³ Lab)!

3/3 🔔 Keep an eye on the repositories and turn on notifications — in the coming days, we’ll release the code and open-access versions of both papers! On to the next! 🚀

2/3 1️⃣ From Scratch to Twin: The Design of Your First Human-Centric Digital Twin 🏆 Best Presentation Award for this! 🥳 🔗 GitHub: https://t.co/p1tXwSFswD 2️⃣ GAIA: A Comprehensive Pipeline for Enabling Aircraft Digital Twin Creation 🔗 GitHub:

1/3 Just came back from an amazing week in #Toronto 🇨🇦, where I had the pleasure of presenting two of our latest works with Francesco Biondani at the @ieeeisie2025! 🏭

4/4 🔝 Tested on renowned industrial datasets such as MVTec‑AD and ViSA, KairosAD delivers top-tier anomaly detection performance — while using 78% fewer parameters and running 4× faster inference times than SOTA alternatives, all without sacrificing AUROC effectiveness.

3/4 📚 Specifically, we leveraged the power of #MobileSAM to build a supervised approach that excels in practical deployment scenarios.

2/4 Key highlights: 🎯 Supervised image-based anomaly detection on par with top models in accuracy. ⚡️ Ultra-efficient: slashes model size and speeds up inference for embedded use. 🌱 KairosAD is also green: its minimal computational footprint ensures low carbon emissions.

1/4 📣 I am super proud of the work we have achieved in #KairosAD — a lightweight, lightning-fast anomaly detection solution tailored for embedded devices! 🚀 🔍 Article, code, and details are available on our website: https://t.co/mxhSpN1zwf 💪

🔍 This repository aims to be your go-to resource to stay updated and deepen your understanding of this fast-evolving field 📱☁️ If you find it useful, please kindly star ⭐ this repository.

3/4 💙 I will actively maintain this repository by incorporating new research as it emerges. If you have any suggestions, find any missed papers, or update any preprint arXiv paper that has been accepted to some venue, please email me or submit a pull request.

2/4 📚 Specifically, this is a curated collection of cutting-edge research, tools, and frameworks focused on #SplitComputing and #EarlyExit, the emerging paradigm that brings learning models closer to the edge by distributing workloads between devices and the cloud.

1/4 🚀 Support the Future of Edge Intelligence! 🧠 Finally, let me share the "Awesome Split Computing" repository! 💡 🔗:

I can bet on the results of the investigation: #NeurIPS, #BMVC, #SIGGRAPH Asia deadline, + #ICCV rebuttals 😅

Honored to be selected as one of the Outstanding Reviewers for #CVPR2025 🥳 I believe that serving the research community with care and responsibility is essential — it's how we help it grow, and how we grow ourselves 🧠 Looking forward to continuing on this path 😎

3/3 Accepted papers will be published in the Springer Book Series 📚 Don't miss the chance to share your research and connect with the AI community! 👉 More info: https://t.co/jwPDlrd5XI #FAIEMA2025 #CallForPapers #AI #Ethics #Research #Conference

2/3 The 3rd International Conference on Frontiers of Artificial Intelligence, Ethics, and Multidisciplinary Applications is calling for contributions on: 🧠 Foundations of AI 🌍 AI-driven Applications ⚖️ Ethics and Governance 🌟 Emerging Multidisciplinary Areas

1/3 🚀 Only 1 month left to submit your paper to FAIEMA 2025! 🚀 🗓 Submission Deadline: 31 May 2025 📍 Join us in beautiful Stavanger, Norway, on 18–19 September! 🇳🇴

4/4 💡Finally, our survey uniquely integrates recent advancements, including diffusion models, Physics-Informed Neural Networks, and foundation models, offering a detailed analysis that sets it apart from previous works. Please, take a look, and let us know what you think! 💪

3/4 📚 We explore major learning models (ConvNets, GANs, Diffusion Models, Transformers, GNNs, and more), various machine learning paradigms (supervised, unsupervised, ensemble, transfer, federated learning), datasets, evaluation metrics, and hardware considerations.

2/4 🔍 In this work, we provide an up-to-date and systematic review of deep learning approaches applied to Predictive Maintenance (PdM) in industrial contexts.