Preprint
Aug 2026
FLEET: Token-Based Feature Extraction for Event Camera-based Reinforcement Learning
This work proposes FLEET (Feature Learning from Events via Efficient Tokenization), a feature extractor that processes event sequences directly and decouples inference cost of the feature extractor's backbone from the sensor's resolution, enabling end-to-end learning without auxiliary losses.
T. Gottwald, Maximilian Schier, Melanie Schaller et al.
· 0 citations