Paper Detail
Leon Arthur Marx, Francesco Barbato, Matteo Caligiuri, Pietro Zanuttigh
Weakly supervised class-incremental semantic segmentation (WILSS) aims to train a segmentation model over multiple steps, each introducing new concepts to be learned with only image-level supervision. We introduce DR.WILSS, an innovative approach to address catastrophic forgetting in continual learning using diffusion-based generative replay. Our framework leverages language clues to guide the diffusion process, employing self-inpainting and regularization techniques to efficiently produce replay data, aiding the learning process. By generating high-quality replay data, the information from previously learned classes can be preserved during continual updates, a critical challenge in incremental learning scenarios. To further align the statistics of replay data with those of training samples, we apply LoRAs to the generative model. Experimental results demonstrate state-of-the-art performance across multiple benchmarks and generative architectures, while avoiding storage of training data and the use of additional resource-demanding tools during training. The proposed technique enables an optimal tradeoff between training complexity and inference-time accuracy, making DR.WILSS a promising solution for real-world applications.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@article{marx2026dr,
title = {DR.WILSS: Diffusion-Based Replay for Weakly Supervised Continual Semantic Segmentation},
author = {Leon Arthur Marx and Francesco Barbato and Matteo Caligiuri and Pietro Zanuttigh},
year = {2026},
abstract = {Weakly supervised class-incremental semantic segmentation (WILSS) aims to train a segmentation model over multiple steps, each introducing new concepts to be learned with only image-level supervision. We introduce DR.WILSS, an innovative approach to address catastrophic forgetting in continual learning using diffusion-based generative replay. Our framework leverages language clues to guide the diffusion process, employing self-inpainting and regularization techniques to efficiently produce repla},
url = {https://arxiv.org/abs/2609.18444},
keywords = {cs.CV},
eprint = {2609.18444},
archiveprefix = {arXiv},
}
{}