EDA Group @ Polito

The EDA Group was at the SystemC AMS & COSEDA User Group Meeting 2024! https://lnkd.in/eS_6W7J8

Congratulations to Francesco Daghero, Matteo Risso and all the other PhD students!

Great work developed in our group, mainly by Matteo Risso, with the help of Marco Castellano and his group in STMicroelectronics

Check out this article by Wired Italia that highlights an innovative project our team is proud to be part of: https://lnkd.in/eS_FvNYJ MASAI is a project funded by the European Space Agency - ESA to support humanitarian efforts in response to the Ukrainian crisis, focusing on the assessment of building damage using satellite images (RADAR and VHR). We, as Politecnico di Torino, contributed to the development of a Deep Learning model which analyses the pre and post-conflict images and generates a building segmentation map with annotated damage. We would like to extend our thanks to all our project partners:  TerrAria s.r.l., Gisky Srl, EUCENTRE and Associazione della Croce Rossa Italiana

On July 18th Chen XIE discussed her Ph.D. Thesis “Synthesis of Smart and Intelligent Sensors”, made in collaboration with STMicroelectronics In her Ph.D. Thesis, she developed efficient machine-learning architectures and optimization techniques to enhance the energy efficiency of microcontroller-based smart sensors at the edge. A huge thanks to the commision: Sara Vinco, Letícia Maria Bolzani Pöhls, Jose Ayala, Donkyu Baek and Francesco Regazzoni and also to her Supervisor: Andrea Calimera Congratulations!

At #ACACES24, our Alessio Burrello is teaching a course on the optimization and compilation of neural networks for #RISCV #multicore processors, including the main steps required to optimize and deploy these models on ultra-low-power embedded hardware like the PULP Platform

Francesco Daghero recently presented our latest paper at the International Conference on Embedded Computer Systems: Architectures, Modeling and Simulation (https://lnkd.in/eFSUwqQY). The paper, titled "Accelerating Depthwise Separable Convolutions on Ultra-Low-Power Devices," is available at https://lnkd.in/ete4FNfV. The paper introduces several edge-tailored layer-fusion approaches for depthwise and pointwise convolutions. With experiments on several state-of-the-art models, we show significant reductions both in terms of memory transfers and latency. It was a fantastic opportunity to showcase our work and enjoy some time by the seaside!

The paper "Integrating SystemC-AMS Power Modeling with a RISC-V ISS for Virtual Prototyping of Battery-operated Embedded Devices" by Mohamed Amine Hamdi, Giovanni Pollo and Matteo Risso is now available on ACM Digital Library. Paper: https://lnkd.in/dJYN5h5K If you want to use it for your research, check out the repository: https://lnkd.in/dJMipGzJ

DUT: Elisa Donati under test 🙋♀️ Real-time gesture recognition at the final hands-on session of track 3 at #EFCLSummerSchool

Today Prof. Daniele Jahier Pagliari and Prof. Alessio Burrello, coordinators of track 3 at #EFCLSummerSchool, did some hands-on session about the deployment of AI applications for biosignal processing on GAP9 platform by GreenWaves Technologies. The whole deployment flow, is done with fully open source tools developed by our research group: PLiNIO: https://lnkd.in/dcNJUWjV MATCH: https://lnkd.in/dHzKr9Ug A great opportunity for all the #EFCLSummerSchool participants to gain some practical experience!