Multi-Domain Incremental Learning for Semantic Segmentation

Garg, Prachi and Saluja, Rohit and Balasubramanian, Vineeth N and Arora, Chetan and Subramanian, Anbumani and Jawahar, C.V. (2022) Multi-Domain Incremental Learning for Semantic Segmentation. In: 22nd IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022, 4 January 2022 through 8 January 2022, Waikoloa.

Text
Proceedings_2022_IEEE_CVF4.pdf - Published Version
Available under License Creative Commons Attribution.
Download (5MB)

Abstract

Recent efforts in multi-domain learning for semantic segmentation attempt to learn multiple geographical datasets in a universal, joint model. A simple fine-tuning experiment performed sequentially on three popular road scene segmentation datasets demonstrates that existing segmentation frameworks fail at incrementally learning on a series of visually disparate geographical domains. When learning a new domain, the model catastrophically forgets previously learned knowledge. In this work, we pose the problem of multi-domain incremental learning for semantic segmentation. Given a model trained on a particular geographical domain, the goal is to (i) incrementally learn a new geographical domain, (ii) while retaining performance on the old domain, (iii) given that the previous domain's dataset is not accessible. We propose a dynamic architecture that assigns universally shared, domain-invariant parameters to capture homogeneous semantic features present in all domains, while dedicated domain-specific parameters learn the statistics of each domain. Our novel optimization strategy helps achieve a good balance between retention of old knowledge (stability) and acquiring new knowledge (plasticity). We demonstrate the effectiveness of our proposed solution on domain incremental settings pertaining to real-world driving scenes from roads of Germany (Cityscapes), the United States (BDD100k), and India (IDD). 1 © 2022 IEEE.

[error in script]

IITH Creators: