Aug 2026· arXiv.org· Vol abs/2608.30281· 0 citations· 62 references
Computer Science
TL;DR
This work proposes SELECT, a novel approach for Selective Context Transfer, which instead grounds each new class in a small set of semantically similar past classes, and adds a controlled noise perturbation and a margin-transfer loss that enforces separation between the new class token and its source tokens.
Abstract
Class-Incremental Semantic Segmentation (CISS) is fundamentally challenged by catastrophic forgetting and background shift, where learning new concepts degrades performance on previously seen classes. While existing methods attempt to balance stability (retaining old knowledge) and plasticity (learning new knowledge), they often fail to leverage prior knowledge effectively. These approaches typically rely on indiscriminate knowledge transfer or ambiguous initializations, which can dilute crucial semantic information. To overcome this limitation, we propose SELECT, a novel approach for Selective Context Transfer, which instead grounds each new class in a small set of semantically similar past classes. Its core is a Context Transfer Attention mechanism that aggregates the learned tokens from similar classes into a structured initialization for the new class. To ensure this transfer does not corrupt the borrowed representations, we add a controlled noise perturbation and a margin-based context-transfer loss that enforces separation between the new class token and its source tokens. Extensive experiments on Pascal VOC and ADE20K show that SELECT consistently outperforms prior work, achieving mIoU of 2.2% on VOC and 2.8% on ADE, providing an effective handle on the stability-plasticity dilemma. Code is available at https://github.com/avigupta2798/SELECT.
This work proposes a framework that curates trustworthy synthetic knowledge for stable G-FSCIL, and designs a boundary-stable incremental adaptation scheme, including synthetic-informed prototype initialization and bidirectional boundary calibration to mitigate old-new conflicts.
Jun-Hui Yin, Yu-Chen Yang, Yi-Lin Yin et al.· Proceedings of the 34th ACM...· 0 citations
Open-vocabulary object detection (OVD) expects the model trained on base classes to recognize both base and novel classes at inference. Typical methods rely on CLIP-style vision-language embeddings, yet the learned space is often biased toward base categories, with limited effective capacity preserved for novel concept...
Wei Zhang, Zhu Teng, Zhixiang Chi et al.· IEEE transactions on multime...· 0 citations
This work proposes a dynamic width expansion method that increases the number of neurons within existing layers according to a normalized loss criterion, without requiring task-specific information.
A. 1. Conde, Y. Elkhatib, C. Ranieri· arXiv.org· 0 citations
Adapting vision-language models to downstream tasks has achieved remarkable success by leveraging pseudo-labels generated from unlabeled data. Existing methods typically assume a uniform unlabeled data distribution, and thus the resulting pseudo-label distribution is likewise uniform. However, real-world data distribut...
Ke-Liang Chen, Ya-Xin Hou, Hui Liu et al.· 0 citations
Semantic segmentation depends heavily on real images with pixel-level annotations, making dense supervision a major obstacle to deployment in new domains. We study task-specific data-free semantic segmentation, where training uses generic pretrained generative priors but has access to no target-domain images, masks, or...
Ming Tian, Marzi Heidari, Yu-Hong Guo· 0 citations
Pretrained vision models provide strong general-purpose representations, but in Domain-Incremental Learning (DIL), where the input distribution changes sequentially, adapting to new domains often degrades performance on previously learned ones. We observe that forgetting is not uniform across domains and classes and is...
Hyunmin Kong, Gyutae Oh, Jitae Shin· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.