Weakly-Supervised Continual Learning for Class-Incremental Segmentation

By Gaston Lenczner, Adrien Chan-Hon-Tong, Nicola Luminari, Bertrand Le Saux. 2022

4 citations recorded by the source corpus.

Transfer learning is a powerful way to adapt existing deep learning models to new emerging use-cases in remote sensing. Starting from a neural network already trained for semantic segmentation, we propose to modify its label space to swiftly adapt it to new classes under weak supervision. To alleviate the background shift and the catastrophic forgetting problems inherent to this form of continual learning, we compare different regularization terms and leverage a pseudo-label strategy. We experimentally show the relevance of our approach on three public remote sensing datasets. Code is open-source and released in this repository: https://github.com/alteia-ai/ICSS.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Valerio Marsocci
2. Simone Scardapane
3. Laura Palagi

In Computer Vision and Pattern Recognition

1. Valerio Marsocci
2. Simone Scardapane
3. Laura Palagi

In Media Technology

1. Simone Scardapane
2. Valerio Marsocci

Citations in the corpus

• Continual Barlow Twins: Continual Self-Supervised Learning for Remote Sensing Semantic Segmentation (2022, background): “other hand, in EO, this combination has not yet been explored, to our knowledge, apart from a few embryonic contributions combining weakly-supervision with CL [11].”