Titouan Cottencin

Papers

Total Citations

H-Index

About

Titouan Cottencin’s research lies at the intersection of computer vision and robotics, with a particular focus on how machines perceive and interpret natural environments. His most cited work, “Evaluation of Off-The-Shelf CNNs for the Representation of Natural Scenes with Large Seasonal Variations” (2017), addresses a critical challenge in autonomous systems: maintaining robust visual recognition across dramatic seasonal changes—from snow-covered landscapes to lush summer foliage. By systematically testing pre-trained convolutional neural networks on outdoor scenes, Cottencin demonstrated that while these models struggle with seasonal shifts, careful feature selection can mitigate performance drops. This contribution has informed subsequent work in long-term visual place recognition for field robotics, where consistent scene understanding is essential for navigation and mapping. Though his citation count (3) reflects a focused, early-career output, the paper’s practical implications for deploying robots in uncontrolled, natural settings underscore its value. Cottencin’s work bridges the gap between off-the-shelf deep learning and real-world environmental variability, offering a foundation for more resilient perception systems in agriculture, forestry, and planetary exploration.