DeepAxe: A Framework for Exploration of Approximation and Reliability Trade-offs in DNN Accelerators

no code implementations • 14 Mar 2023 • Mahdi Taheri, Mohammad Riazati, Mohammad Hasan Ahmadilivani, Maksim Jenihhin, Masoud Daneshtalab, Jaan Raik, Mikael Sjodin, Bjorn Lisper

The framework enables selective approximation of reliability-critical DNNs, providing a set of Pareto-optimal DNN implementation design space points for the target resource utilization requirements.

TAS: Ternarized Neural Architecture Search for Resource-Constrained Edge Devices

1 code implementation • Design, Automation and Test in Europe Conference (DATE) 2022 • Mohammad Loni, Hamid Mousavi, Mohammad Riazati, Masoud Daneshtalab, and Mikael Sjodin

This paper proposes TAS, a framework that drastically reduces the accuracy gap between TNNs and their full-precision counterparts by integrating quantization into the network design.

Neural Architecture Search Quantization