Manish Kansana

Papers

Total Citations

H-Index

About

Manish Kansana is a researcher at the forefront of multimodal perception for robotics, with a primary focus on integrating tactile and visual data to enhance machine understanding of physical environments. His most notable contribution is the development of Surformer v1, a pioneering transformer-based architecture for surface classification that fuses structured tactile features with visual inputs. This work, published in 2025 and already garnering 2 citations, addresses a critical challenge in robotic interaction: enabling machines to recognize and differentiate materials by touch and sight simultaneously. By leveraging attention mechanisms, Kansana’s approach moves beyond traditional single-modality methods, offering a more robust and context-aware solution for tasks ranging from autonomous manipulation to assistive robotics. His research sits at the intersection of deep learning, sensor fusion, and embodied AI, with clear implications for improving how robots perceive and interact with their surroundings. Kansana’s work is particularly significant for students and researchers exploring transformer-based models in robotics, as it demonstrates a practical pathway for combining disparate sensory streams into a unified perceptual system. With a growing citation footprint and a focus on real-world applicability, he is establishing himself as a key voice in the evolution of tactile-aware artificial intelligence.