Akash Dhamasia
Papers
1
Total Citations
5
H-Index
1
About
Akash Dhamasia is a researcher at the forefront of safety-critical artificial intelligence, specializing in the dependability of deep learning systems deployed in high-stakes environments. His work focuses on ensuring the robustness of convolutional neural networks (CNNs) against hardware-induced faults, a critical challenge for applications in autonomous driving and human-robot interaction. Dhamasia’s most cited paper, "Towards a Safety Case for Hardware Fault Tolerance in Convolutional Neural Networks Using Activation Range Supervision" (2021), introduces a novel framework for detecting and mitigating soft errors that corrupt neural network computations. By leveraging activation range supervision, his approach provides a formal safety case for hardware fault tolerance, bridging the gap between theoretical AI reliability and real-world deployment. This work has garnered attention for its practical implications in certifying AI systems for safety-critical use. Dhamasia’s contributions are essential reading for researchers and engineers working to make deep learning trustworthy in applications where failure is not an option, establishing him as a key voice in the emerging field of dependable AI systems.
Research Focus
Key Achievements
Top Papers
- 1