In many applications, pinpoint accuracy in keypoint detection is paramount. What if we could achieve this on low-compute microcontrollers? 🤯 Introducing a groundbreaking approach! 🚀 ...optimizing NN for embedded systems, ensuring efficiency without sacrificing accuracy. 💪

🚨 Exciting Opportunity Alert! 🚨 Passionate about #tinyML & distributed intelligence in #IoT? Fondazione Bruno Kessler - FBK's E3DA unit & UniTrento_DISI Doctoral School are offering a PhD scholarship on Artificial Intelligence for Tiny, Connected Devices - A5 Details: iecs.unitn.it/education/admi…

📑Our publication "XimSwap: Many-to-Many Face Swapping for TinyML | ACM Transactions on Embedded Computing Systems" is now is now available in the ACM Digital Library🚀🚀🚀 dl.acm.org/doi/10.1145/36… francesco paissan Alberto Ancilotto Fondazione Bruno Kessler - FBK

next week, I will present "tinyCLAP👏: distilling constrastive language-audio pretrained models" at the tinyML Innovation Forum check out the code and play with the model! 🤗demo: shorturl.at/Yf3sT 🖥️code: shorturl.at/hu2gL project webpage: shorturl.at/drbRM

So excited to see "tinyCLAP👏" presented next week at the tinyML EMEA Innovation Forum! tinyML Foundation It was an amazing work that will go far. Don't miss out—check out the demo and explore our code! 🤗🖥️🚀 #tinyML #AI #audioprocessing #embedded Fondazione Bruno Kessler - FBK

🥳Congrats to Alessandro! 🚀 Check out this new study that is one chapter of a broader study on the effects of IMU sensor-to-segment calibration on different segments of the upper limb (wrist, elbow, shoulder). A must-read for anyone in biomechanics! 📚👏

⏳⌛Last day at tinyML EMEA Forum in Milan. Come to discuss with Alberto Ancilotto about how generative AI can fit in a microcontroller 🚀🚀🚀

🎉 Foundation models on tiny devices? francesco paissan explained how at the #tinyML EMEA Innovation Forum 2024 in Milan presenting "tinyCLAP: distilling language-audio pretrained models"🚀 In September, this work will be presented INTERSPEECH 2025. Stay tuned! E3DA Research Unit at DigiS center, FBK Fondazione Bruno Kessler - FBK

Alessandro is presenting now at IEEE COINS 2024 coinsconf.com

🚀 Partner Spotlight: FBK 🚀 Fondazione Bruno Kessler - FBK drives data anonymisation and privacy in the project. They promote innovation and tech development. FBK also works on knowledge sharing, commercialisation, and IP management. Discover FBK’s role: patternproject.eu/partners/ #PATTERNProject

Two more days until our #conference starts! We look forward to welcoming our participants at the UMCG in Groningen, for 3 days of exciting presentations and interactions. Our final program and book of abstracts is now online! Check it out here: repairs-etn.eu/closing-confer…

✨"Congrats to Stefano Ciapponi for presenting his first publication as PhD student, 'tinyVocos: Neural Vocoders on MCUs,' at the Internet of Sounds Symposium! 🎉 An exciting step forward in tinyML for sound processing. 🚀👏 Fondazione Bruno Kessler - FBK 🔗 internetofsounds.net/is2_2024/"

🚀 Excited to share that Alberto Ancilotto will present "Neural Style Transfer on STM32 MCUs" at the GenAI tinyML Foundation event next Oct 17-18, 2024 🎨✨ Learn how we're enabling GenAI art & video anonymization on low-power devices with XiNet for edge computing. Fondazione Bruno Kessler - FBK

Our research was at #ECCV2024 FOMO Workshop. Elisabetta Farella presented "Unlocking the potential of Foundation Models in tinyML," about tinyCLAP and tiny Neural Style transfer exploiting FM power! 👏 Thanks to the organizing team! Let’s drive AI towards a sustainable future! 🚀

If you want to know more, have a look at the research of francesco paissan at Interspeech 2024 and of Alberto Ancilotto at PeRConAI 2024 Special thanks to Yiming Wang 🙏, for bringing this important topic to the forefront. Let's continue driving AI towards a more sustainable future! 🚀

Alberto Alberto Ancilotto is presenting now at the tinyML virtual event on Generative AI at the edge! Hurry up and connect to attend his presentation! tinyML Foundation Fondazione Bruno Kessler - FBK

🚀 Thrilled !! Our paper "Linear Transformers beat YOLO for Embedded Object Detection" has been accepted to PeRConAI 2025! 🎉 We bring ViT-level performance to tiny devices with faster, smaller models. Stay tuned!👂#tinyML #embeddedAI Alberto Ancilotto Edoardo Castagnini Elisabetta Farella