# ModuSeg: Decoupling Object Discovery and Semantic Retrieval for Training-Free Weakly Supervised Segmentation Source: [ModuSeg: Decoupling Object Discovery and Semantic Retrieval for Training-Free Weakly Supervised Segmentation]() Collection index: [llms.txt]() Data updated: 2026-07-01T05:00:18.486Z Authors: Qingze He, Fagui Liu, Dengke Zhang, Qingmao Wei, Quan Tang Published: 2026-04-08T12:38:07.000Z ## Abstract Weakly supervised semantic segmentation aims to achieve pixel-level predictions using image-level labels. Existing methods typically entangle semantic recognition and object localization, which often leads models to focus exclusively on sparse discriminative regions. Although foundation models show immense potential, many approaches still follow the tightly coupled optimization paradigm, struggling to effectively alleviate pseudo-label noise and often relying on time-consuming multi-stage retraining or unstable end-to-end joint optimization. To address the above challenges, we present ModuSeg, a training-free weakly supervised semantic segmentation framework centered on explicitly decoupling object discovery and semantic assignment. Specifically, we integrate a general mask proposer to extract geometric proposals with reliable boundaries, while leveraging semantic foundation models to construct an offline feature bank, transforming segmentation into a non-parametric feature retrieval process. Furthermore, we propose semantic boundary purification and soft-masked feature aggregation strategies to effectively mitigate boundary ambiguity and quantization errors, thereby extracting high-quality category prototypes. Extensive experiments demonstrate that the proposed decoupled architecture better preserves fine boundaries without parameter fine-tuning and achieves highly competitive performance on standard benchmark datasets. Code is available at https://github.com/Autumnair007/ModuSeg. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Semantic Segmentation on COCO 2014 (val)](): Semantic segmentation; COCO 2014 (val); mIoU: 56.7. - [Semantic Segmentation on PASCAL VOC 2012 (val)](): Semantic segmentation; PASCAL VOC 2012 (val); Mean IoU: 86.3. - [Semantic Segmentation on PASCAL VOC (val)](): Semantic segmentation; PASCAL VOC (val); mIoU: 86.3. - [Semantic Segmentation on PASCAL VOC 2012 (test)](): Semantic segmentation; PASCAL VOC 2012 (test); mIoU: 86.6. ## Original sources and implementation [Original paper]() [PDF]()