Alex Dimakis
Papers
1
Total Citations
1
H-Index
1
About
Alex Dimakis is a leading researcher in computer vision and representation learning, with a focus on bridging extreme viewpoint gaps between egocentric and exocentric video data. His most notable contribution, the VIEWPOINTROSETTA framework, introduces a novel approach to learning viewpoint-invariant representations by leveraging large-scale unpaired ego- and exo-centric videos. This work, published in 2025, addresses a critical challenge in augmented reality and robotics—enabling systems to understand human actions regardless of camera perspective. Though early in its citation trajectory, the paper’s innovative methodology has already garnered attention for its potential to unlock new capabilities in cross-view action recognition. Dimakis’s research sits at the intersection of self-supervised learning, multi-modal perception, and embodied AI, pushing the boundaries of how machines interpret human behavior from diverse visual inputs. His work promises to transform applications from assistive robotics to immersive AR experiences, making him a rising voice in the field of view-invariant representation learning.
Research Focus
Key Achievements
Top Papers
- 1