S. Narayan

Papers

Total Citations

H-Index

About

S. Narayan’s research centers on hardware implementations of neural networks for embedded pattern recognition, with a particular emphasis on achieving rotation invariance—a critical challenge for real-world vision systems. Their most-cited work, “A rotation-invariant embedded pattern recognition system” (2003, 3 citations), tackles the performance gap between software and hardware neural networks by proposing architectural designs that exploit inherent parallelism. This contribution addresses the longstanding limitation that hardware neural networks, while faster, are often too costly for low‑cost pattern recognition systems. By demonstrating how rotation invariance can be embedded directly into hardware, Narayan’s work provides a pathway toward affordable, high‑performance recognition in resource‑constrained environments. Though citation counts are modest, the research is notable for its practical focus on bridging algorithmic robustness with efficient hardware implementation—a topic of growing relevance as edge computing and embedded AI expand. Narayan’s contributions offer valuable insights for students and engineers seeking to design neural network accelerators that are both rotation‑tolerant and cost‑effective.