Abstract
As one of the most important underwater sensing technologies, forward-looking sonar exhibits unique imaging characteristics. Sonar images are often affected by severe speckle noise, low texture contrast, acoustic shadows, and geometric distortions. These factors make it difficult for traditional teacher-student frameworks to achieve satisfactory performance in sonar semantic segmentation tasks under extremely limited labeled data conditions. To address this issue, we propose a Collaborative Teacher Semantic Segmentation Framework for forward-looking sonar images. This framework introduces a multi-teacher collaborative mechanism composed of one general teacher and multiple sonar-specific teachers. By adopting a multi-teacher alternating guidance strategy, the student model can learn general semantic representations while simultaneously capturing the unique characteristics of sonar images, thereby achieving more comprehensive and robust feature modeling. Considering the challenges of sonar images, which can lead teachers to generate a large number of noisy pseudo-labels, we further design a cross-teacher reliability assessment mechanism. This mechanism dynamically quantifies the reliability of pseudo-labels by evaluating the consistency and stability of predictions across multiple views and multiple teachers, thereby mitigating the negative impact caused by noisy pseudo-labels. Notably, on the FLSMD dataset, when only 2% of the data is labeled, our method achieves a 5.08% improvement in mIoU compared to other state-of-the-art approaches.
Abstract (translated)
作为最重要的水下感知技术之一，前视声呐具有独特的成像特性。声呐图像常受严重散斑噪声、低纹理对比度、声影及几何畸变的影响。这些因素导致传统师生框架在标注数据极度有限的条件下难以在声呐语义分割任务中取得理想性能。针对这一问题，我们提出了一种面向前视声呐图像的协同教师语义分割框架。该框架引入由一位通用教师与多位声呐专用教师构成的多教师协同机制，通过采用多教师交替引导策略，使学生模型既能学习通用语义表征，又能同时捕捉声呐图像的独特特征，从而实现更全面稳健的特征建模。考虑到声呐图像的挑战可能导致教师生成大量噪声伪标签，我们进一步设计了跨教师可靠性评估机制。该机制通过评估多视图及多教师间预测的一致性与稳定性，动态量化伪标签的可靠性，从而缓解噪声伪标签带来的负面影响。值得注意的是，在FLSMD数据集上，当仅使用2%标注数据时，本方法在mIoU指标上相较于其他先进方法取得了5.08%的性能提升。
URL
https://arxiv.org/abs/2603.21071
