Wilfred Gomes
Papers
1
Total Citations
32
H-Index
1
About
Wilfred Gomes is a rising leader at the intersection of hardware-software co-design and energy-efficient artificial intelligence, with a core focus on compute-in-memory (CIM) architectures for edge intelligence. His most-cited work, "MC-CIM: Compute-in-Memory With Monte-Carlo Dropouts for Bayesian Edge Intelligence" (2022, 32 citations), pioneers a novel framework that integrates Monte Carlo dropout sampling directly into CIM hardware. This breakthrough enables deep neural networks to express prediction uncertainties—a critical capability for high-stakes applications like autonomous driving or medical diagnostics—while maintaining the ultra-low power consumption essential for edge devices. By addressing the fundamental limitation of deterministic DNNs, which cannot quantify their own confidence, Gomes’s research bridges Bayesian deep learning with practical hardware acceleration. His work stands out for tackling both algorithmic robustness and hardware efficiency, offering a path toward trustworthy AI deployment in resource-constrained environments. With growing recognition in the circuits and systems community, Gomes is shaping the future of Bayesian edge intelligence, where reliable, uncertainty-aware computation meets the stringent power budgets of real-world sensors and IoT devices.
Research Focus
Key Achievements
Top Papers
- 1