Jacek Tabor

Papers

Total Citations

H-Index

About

Jacek Tabor is a computer vision researcher whose work tackles two critical challenges in modern AI: handling missing data and generating efficient 3D object representations. His most notable contribution, "MisConv: Convolutional Neural Networks for Missing Data" (2022, 7 citations), addresses the fundamental problem of processing incomplete data in CNNs—a challenge that arises in practical applications from image inpainting to autonomous vehicle navigation. Rather than relying on traditional imputation-based techniques, Tabor's approach offers a more robust solution for real-world scenarios where sensor data is often incomplete. In parallel, his work "HyperFlow: Representing 3D Objects as Surfaces" (2020, 5 citations) introduces a novel generative model that uses hypernetworks to create continuous, lightweight 3D surface representations directly from point clouds. This innovation has significant implications for computer vision applications requiring efficient object representations, such as robotics and augmented reality. Tabor's research demonstrates a consistent focus on developing practical, deployable solutions for fundamental computer vision challenges, bridging the gap between theoretical advances and real-world applications.