Sign in | Register to Bookmark

Enhanced Feature Fusion for Thermal-Color Semantic Segmentation

In this talk we present a new approach for feature fusion between RGB and LWIR Thermal images for the task of semantic segmentation for driving perception. We propose a double DeepLab architecture with specialized encoder-decoders for thermal and color modalities and a shared decoder for final segmentation. We combine two strategies for feature fusion: confidence weighting and correlation weighting. We report state-of-the-art on the thermal-color semantic segmentation task.

Want to watch this session?

You must be logged in and have an AutoSens Brussels 2022 pass to watch this session.

Please either login or create an account.

Already registered? Please Log in

Hear from:


Oriel Frigo
AI Engineer

Scroll to Top