Prototype-based classification is a classical method in machine learning, and recently it has achieved remarkable success in semi-supervised semantic segmentation. However, the current approach isolates the prototype initialization process from the main training framework, which appears to be unnecessary. Furthermore, while the direct use of K-Means algorithm for prototype generation has considered rich intra-class variance, it may not be the optimal solution for the classification task. To tackle these problems, we propose a novel boundary-refined prototype generation (BRPG) method, which is incorporated into the whole training framework. Specifically, our approach samples and clusters high- and low-confidence features separately based on a confidence threshold, aiming to generate prototypes closer to the class boundaries. Moreover, an adaptive prototype optimization strategy is introduced to make prototype augmentation for categories with scattered feature distributions. Extensive experiments on the PASCAL VOC 2012 and Cityscapes datasets demonstrate the superiority and scalability of the proposed method, outperforming the current state-of-the-art approaches. The code is available at xxxxxxxxxxxxxx.

翻译：基于原型的分类是机器学习中的经典方法，近年来在半监督语义分割领域取得了显著成功。然而，当前方法将原型初始化过程与主训练框架相隔离，这一做法显得不必要。此外，尽管直接使用K-Means算法生成原型时考虑了丰富的类内方差，但这可能并非分类任务的最优解决方案。为解决这些问题，我们提出了一种新颖的边界细化原型生成（BRPG）方法，并将其融入整体训练框架中。具体而言，我们的方法基于置信度阈值分别对高置信度和低置信度特征进行采样和聚类，旨在生成更接近类别边界的原型。此外，我们引入了一种自适应原型优化策略，为特征分布分散的类别进行原型增强。在PASCAL VOC 2012和Cityscapes数据集上的大量实验表明，所提方法具有优越性和可扩展性，性能超越了当前最先进的算法。代码已开源至 xxxxxxxxxxxxxx。