Pablo F. Alcantarilla
Toshiba (Japan), University of Clermont Auvergne, Concentration Heat and Momentum (United Kingdom)
Papers
7
Total Citations
258
H-Index
6
About
Pablo F. Alcantarilla is a leading researcher in robotics and computer vision, specializing in life-long visual localization—a critical challenge for autonomous systems operating in dynamic environments. His work addresses the extreme appearance variability caused by seasons, illumination, and weather, which has long hindered robust topological localization. Alcantarilla’s most influential contributions include the development of ABLE-M, a method that efficiently matches binary sequences from images to enable reliable place recognition across time. His 2016 paper on fusing and binarizing CNN features for robust topological localization across seasons has garnered 99 citations, while his 2015 work on life-long localization using ABLE-M has 96 citations, underscoring their impact on the field. Beyond seasonal localization, Alcantarilla has advanced vision-based localization for humanoid robots and visually impaired individuals, demonstrating the broader applicability of his research. His recent work on real-time semantic mapping with latent prior networks (SeMLaPS, 2023) integrates 2D and 3D neural networks to enhance SLAM systems, pushing the boundaries of robotic perception. With over 250 total citations, Alcantarilla’s contributions are foundational for life-long autonomy in robotics and intelligent vehicles.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Are you ABLE to perform a life-long visual topological localization?22 citations · 2017
- 4How to localize humanoids with a single camera?18 citations · 2012
- 5Vision based localization: from humanoid robots to visually impaired people11 citations · 2011
- 6
- 7