Ashish Shah

Papers

Total Citations

H-Index

About

Ashish Shah is a researcher specializing in computer vision and generative artificial intelligence, with a particular focus on unsupervised motion transfer and image animation. His most-cited work, "Self-appearance-aided Differential Evolution for Motion Transfer" (2021), addresses a key challenge in the field: transferring the motion of a driving video to a static object in a source image while preserving the source's identity, without relying on labeled data or domain priors. This contribution advances the state of the art in unsupervised motion transfer, offering a novel approach that leverages self-appearance features to improve fidelity and robustness. Although his citation count is currently modest, his research lays important groundwork for applications in animation, video editing, and virtual avatars. Shah’s work is notable for tackling the difficult problem of maintaining identity consistency during motion transfer, a critical step toward more realistic and practical generative models. His findings are particularly relevant for researchers exploring differential evolution techniques in deep learning, and his contributions continue to inspire further innovation in unsupervised video-to-video synthesis.