Michalis Rizakis
Papers
1
Total Citations
6
H-Index
1
About
Michalis Rizakis is a researcher at the forefront of efficient deep learning hardware, specializing in approximate computing and FPGA-based acceleration for neural networks. His work addresses the critical challenge of deploying complex models like LSTMs under strict computation time constraints, a key concern for real-time and embedded systems. In his most-cited paper, "Approximate FPGA-Based LSTMs Under Computation Time Constraints" (2018), Rizakis pioneered techniques that trade off minor accuracy for significant gains in processing speed and energy efficiency, enabling LSTM inference on resource-limited devices. This contribution has garnered 6 citations, reflecting its relevance to the growing field of edge AI. Rizakis’s research bridges the gap between algorithmic innovation and practical hardware implementation, offering scalable solutions for time-sensitive applications such as autonomous systems and IoT. His work stands as a testament to the power of approximate computing in pushing the boundaries of what is achievable with constrained computational resources, making him a notable figure in the intersection of machine learning and reconfigurable hardware design.
Research Focus
Key Achievements
Top Papers
- 1Approximate FPGA-Based LSTMs Under Computation Time Constraints6 citations · 2018