Yash Pratap Singh

About

Yash Pratap Singh is a rising researcher in artificial intelligence and human activity recognition, with a focus on spatiotemporal modeling and sensor-based perception. His most-cited work introduces a **Sparse and Contractive Graph-Based Variational Encoder-Decoder with Multihead Attention**—a novel architecture designed to robustly capture complex spatiotemporal dynamics from sensor data. This contribution addresses critical limitations in existing data-driven methods, particularly in domains like person surveillance and human-robot interaction, where accurate and resilient activity recognition is essential. By integrating graph-based variational inference with multihead attention mechanisms, Singh’s approach achieves enhanced representational efficiency and robustness against noisy or incomplete inputs. Although early in his career, his work has already garnered attention for its technical depth and practical relevance, signaling a promising trajectory in advancing human-centric AI systems. His research sits at the intersection of deep learning, graph neural networks, and embodied perception, offering impactful solutions for real-world sensing challenges.

Research Focus

Key Achievements

Top Papers

1. 1
   Sparse and Contractive Graph-Based Variational Encoder-Decoder with Multihead Attention for Robust Spatiotemporal Activity Recognition

   1 citations · 2025

Key Collaborators

Contact & Links