Hampus Linander

Papers

Total Citations

H-Index

About

Hampus Linander is a researcher at the forefront of geometric deep learning and computer vision, with a particular focus on spherical and omnidirectional data. His most notable contribution is the development of **HEAL-SWIN**, a Vision Transformer architecture designed to operate directly on the sphere, introduced in his 2024 paper. This work addresses a critical challenge in robotics and autonomous driving: the severe distortion and information loss that occurs when standard neural networks process wide-angle fisheye images through traditional planar projections. By leveraging the HEALPix spherical discretization, Linander’s approach enables high-resolution, distortion-aware processing of 360° visual data. Though his work is still emerging, with his flagship paper already garnering 8 citations, it has quickly captured the attention of the autonomous systems community for its potential to improve perception in navigation and scene understanding. Linander’s research bridges the gap between spherical geometry and modern transformer architectures, offering a principled alternative to conventional projection-based methods. His contributions are particularly relevant for applications requiring robust, wide-field-of-view perception, positioning him as a promising young voice in the intersection of geometry, vision, and robotics.