Abstract:Weakly-Supervised Semantic Segmentation (WSSS) aims to train segmentation models by weak labels, which is receiving significant attention due to its low annotation cost. Existing approaches focus on generating pseudo labels for supervision while largely ignoring to leverage the inherent semantic correlation among different pseudo labels. We observe that pseudo-labeled pixels that are close to each other in the feature space are more likely to share the same class, and those closer to the distribution centers tend to have higher confidence. Motivated by this, we propose to model the underlying label distributions and employ cross-label constraints to generate more accurate pseudo labels. In this paper, we develop a unified WSSS framework named Adaptive Gaussian Mixtures Model, which leverages a GMM to model the label distributions. Specifically, we calculate the feature distribution centers of pseudo-labeled pixels and build the GMM by measuring the distance between the centers and each pseudo-labeled pixel. Then, we introduce an Online Expectation-Maximization (OEM) algorithm and a novel maximization loss to optimize the GMM adaptively, aiming to learn more discriminative decision boundaries between different class-wise Gaussian mixtures. Based on the label distributions, we leverage the GMM to generate high-quality pseudo labels for more reliable supervision. Our framework is capable of solving different forms of weak labels: image-level labels, points, scribbles, blocks, and bounding-boxes. Extensive experiments on PASCAL, COCO, Cityscapes, and ADE20K datasets demonstrate that our framework can effectively provide more reliable supervision and outperform the state-of-the-art methods under all settings. Code will be available at

Enhanced Pseudo-Label Generation with Self-supervised Training for Weakly-supervised Semantic Segmentation

Learning Pseudo Labels for Semi-and-weakly Supervised Semantic Segmentation

Modeling the Label Distributions for Weakly-Supervised Semantic Segmentation

Self-Supervised Equivariant Attention Mechanism for Weakly Supervised Semantic Segmentation

Image Understands Point Cloud: Weakly Supervised 3D Semantic Segmentation via Association Learning

Decoupled Spatial Neural Attention for Weakly Supervised Semantic Segmentation.

Superpixel Consistency Saliency Map Generation for Weakly Supervised Semantic Segmentation of Remote Sensing Images

Fully Data-Driven Pseudo Label Estimation for Pointly-Supervised Panoptic Segmentation

Weakly Supervised 3D Segmentation Via Receptive-Driven Pseudo Label Consistency and Structural Consistency.

Collaborative Semantic Segmentation with Image Labels.

Progressive Feature Self-reinforcement for Weakly Supervised Semantic Segmentation

Segment Anything is A Good Pseudo-label Generator for Weakly Supervised Semantic Segmentation

Adaptive Affinity Loss and Erroneous Pseudo-Label Refinement for Weakly Supervised Semantic Segmentation

Saliency Guided Self-Attention Network for Weakly and Semi-Supervised Semantic Segmentation

Enhancing Weakly Supervised Semantic Segmentation with Multi-modal Foundation Models: An End-to-End Approach

Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels

Weakly supervised semantic segmentation based on superpixel affinity

Weakly Supervised Semantic Segmentation with Boundary Exploration

Learning from Weak and Noisy Labels for Semantic Segmentation.

Weakly-Supervised Semantic Segmentation with Image-Level Labels: from Traditional Models to Foundation Models

Multi-Granularity Denoising and Bidirectional Alignment for Weakly Supervised Semantic Segmentation