<\/i> Workshop Description<\/h3>\n
<\/p>\n
- <\/p>\n
- Efficient on-device foundation models for vision, audio, and multimodal tasks<\/li>\n
<\/p>\n
- Aggressive quantization, pruning, and neural architecture search tailored to wearable constraints<\/li>\n
<\/p>\n
- HW\/SW co-design and novel accelerators for always-on sensing<\/li>\n
<\/p>\n
- Three-tier architectures combining on-device, on-companion (phone\/edge), and cloud inference<\/li>\n
<\/p>\n
- Speech enhancement and hearing assistance on glasses<\/li>\n
<\/p>\n
- Gaze-, IMU-, and context-aware adaptation<\/li>\n
<\/p>\n
- Privacy-preserving and personalized on-device learning<\/li>\n
<\/p>\n
- Benchmarks and datasets for glasses-mounted sensing<\/li>\n
<\/ul>\n
The workshop extends beyond the main conference tracks by focusing on the tightly coupled stack (from silicon to model to user experience) specific to eyewear form factors. It aims to connect a mature academic community in efficient and multimodal AI with a fast-growing industrial community building the next generation of intelligent glasses.<\/pee><\/div>\n <\/div>\n
<\/p>\n
<\/p>\n<\/i> Submission Details<\/h3>\n
<\/p>\n
<\/p>\n- <\/p>\n
- Format:<\/strong> 4-page short paper, IEEE conference template (excluding references)<\/li>\n
<\/p>\n
- Review:<\/strong> Double-blind, via OpenReview. Select the “Workshop Track” during submission.<\/li>\n
<\/p>\n
- Proceedings:<\/strong> Accepted full papers will appear in the IEEE AIMLSystems 2026 proceedings.<\/li>\n
<\/p>\n
- Other Formats:<\/strong> Non-archival 2-page extended abstracts and live-demo proposals also welcome (presented at the poster\/demo session; not included in proceedings)<\/li>\n
<\/ul>\n
<\/div>\n
<\/div>\n
<\/p>\n
<\/p>\n<\/i> Tentative Schedule<\/h3>\n
<\/p>\n
<\/p>\n<\/p>\nTime<\/div>\n<\/p>\n
Event<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Welcome and Opening Remarks<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Opening Keynote:<\/strong> Academic Perspective on Efficient Multimodal Foundation Models for Wearable AI<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Regular Paper Session I<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Tech Talk \/ Industry Session<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Poster and Live-Demo Session<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
<\/i> Lunch Break<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Afternoon Keynote:<\/strong> Industrial Perspective on Deploying AI at the Glasses Edge<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Regular Paper Session II<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Panel Discussion:<\/strong> “The Glasses-Edge Stack: What’s Missing from Today’s AI-ML Systems Research?”<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n[TBD]<\/div>\n<\/p>\n
Closing Remarks<\/div>\n<\/div>\n
<\/div>\n
<\/div>\n
<\/p>\n
<\/p>\n<\/i> Workshop Chairs<\/h3>\n
<\/p>\n
<\/p>\n<\/p>\n<\/p>\n<\/i><\/div>\n<\/p>\n
<\/p>\nManuel Roveri<\/h4>\n
DEIB, Politecnico di Milano, Italy<\/pee><\/div>\n <\/div>\n
<\/p>\n
Manuel Roveri is Full Professor at DEIB, Politecnico di Milano. His research focuses on Tiny Machine Learning, embedded and edge AI, IoT, and privacy-preserving machine learning. He is a Senior Member of IEEE, Associate Editor of several IEEE Transactions (AI, Neural Networks and Learning Systems, Emerging Topics in Computational Intelligence). He served as General Co-chair of AIMLSystems 2024 and received the 2021 IEEE TETCI Outstanding Associate Editor award. He has published over 100 papers, holds patents in edge AI, and his group collaborates with the Smart Eyewear Lab on on-device wearable intelligence.<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n<\/p>\n<\/i><\/div>\n<\/p>\n
<\/p>\nFabrizio Pittorino<\/h4>\n
DEIB, Politecnico di Milano, Italy<\/pee><\/div>\n <\/div>\n
<\/p>\n
Fabrizio Pittorino is Assistant Professor at DEIB, Politecnico di Milano, and a member of the AI-Tech Research Lab. His research spans efficient and adaptive AI (quantization, binary neural networks, edge and embedded AI) and the theoretical foundations of deep learning through loss-landscape geometry and statistical physics. He is co-inventor on a European patent application on dynamic nested quantization filed with the Politecnico di Milano – EssilorLuxottica Smart Eyewear Lab, and Co-PI on an Infineon-funded project on TinyML. He served as Position Paper Vice-Chair of IJCNN 2026.<\/div>\n<\/div>\n
<\/p>\n
<\/p>\n<\/p>\n<\/i><\/div>\n<\/p>\n
<\/p>\nManuel Pariente<\/h4>\n
EssilorLuxottica, Smart Eyewear R&D, France<\/pee><\/div>\n <\/div>\n
<\/p>\n
Manuel Pariente is AI Research Lead for smart eyewear audio intelligence at EssilorLuxottica. He received his PhD from Universit\u00e9 de Lorraine \/ Inria in 2021 on phase modeling in deep learning-based source separation. He is the creator of Asteroid, a widely-used open-source PyTorch toolkit for audio source separation. In 2022, he co-founded Pulse Audition to build the first AI-powered hearing glasses, acquired by EssilorLuxottica in January 2025. He now leads research on real-time speech enhancement and multimodal audio understanding running on smart eyewear under strict power budgets.<\/div>\n<\/div>\n
<\/div>\n
<\/div>\n
<\/div>\n
<\/p>\n
- Review:<\/strong> Double-blind, via OpenReview. Select the “Workshop Track” during submission.<\/li>\n
- Format:<\/strong> 4-page short paper, IEEE conference template (excluding references)<\/li>\n