Chaitanya Ghone

Papers

Total Citations

H-Index

About

Chaitanya Ghone is a researcher whose work bridges deep learning and signal processing, with a primary focus on developing efficient convolutional neural network (CNN) architectures. His most-cited paper, "Efficient frequency domain CNN algorithm" (2017), introduces a novel approach to accelerating CNN computations by leveraging frequency domain transformations, addressing the computational bottlenecks that limit the deployment of deep learning models in resource-constrained environments such as automotive, industrial, and medical applications. This work, which has garnered 2 citations, proposes a method to reduce the complexity of 2D convolutions, non-linearity, and spatial pooling layers—core components of modern CNNs—by performing operations in the frequency domain rather than the spatial domain. Ghone’s contribution is particularly significant for real-time image classification tasks where latency and power efficiency are critical. By optimizing the trade-off between accuracy and computational cost, his research has implications for advancing edge AI and embedded vision systems. Though early in his career, Ghone’s work demonstrates a clear commitment to making deep learning more practical and scalable, offering a foundation for future innovations in efficient neural network design.