Martin Matak

Papers

Total Citations

H-Index

About

Martin Matak is a robotics researcher specializing in dexterous robotic manipulation, with a particular focus on grasp planning and multimodal sensory integration. His work addresses one of the most persistent challenges in robotics: enabling robotic hands to reliably grasp and manipulate objects without precise prior knowledge of their geometry. His most notable contribution, "Planning Visual-Tactile Precision Grasps via Complementary Use of Vision and Touch" (2023, 22 citations), demonstrates how combining visual and tactile sensing can significantly improve fingertip grasp planning for multi-fingered robotic hands — a capability critical for advanced tasks like tool use, object insertion, and in-hand manipulation. This research is particularly impactful in scenarios where accurate object models are unavailable, a common real-world limitation. Matak's earlier work, "Learning Continuous 3D Reconstructions for Geometrically Aware Grasping" (2020, 7 citations), pushed the boundaries of deep learning-based grasp synthesis by enabling robots to explicitly reason about full 3D object geometry from partial views, moving beyond indirect geometric inference that characterised prior approaches. Together, his contributions reflect a coherent research vision: equipping robots with richer perceptual understanding to achieve more robust and generalizable manipulation, making his work highly relevant to researchers advancing robot autonomy and embodied intelligence.