Best demo paper award

The VLSI Symposia is an international conference on semiconductor technology and circuits that covers everything from process technology to systems-on-chip. Its 2021 edition was held virtually on June 13-19, combining workshops, talks, technical sessions, and live demonstrations.

Petar Jokic's co-authors are: Erfan Azarkhish, CSEM, Regis Cattenoz, CSEM, Luca Benini, ETH Zurich, and Stephane Emery, CSEM.

Results in brief

Smart vision-based IoT applications operate on a sub-mW power budget while requiring power-hungry always-on image processing capabilities. The awarded paper presents a system-on-chip (SoC) that enables hierarchical processing of face analysis under multiple sub-mW operating scenarios using two tightly coupled machine learning (ML) accelerators. A dynamically scalable binary decision tree (BDT) engine for face detection (FD) allows the triggering of a multi-precision convolutional neural network (CNN) engine for subsequent face recognition (FR). The 22nm SoC can therefore dynamically trade-off image analysis depth, frames-per-second (FPS), accuracy, and power consumption. It implements complete end-to-end edge processing, enabling always-on FD and FR within the tight 1mW power budget of a 55mm diameter indoor solar panel.

More information about the the VLSI Symposia: www.vlsisymposium.org