Beyond Pixels: Semi-supervised Semantic Segmentation with a Multi-scale Patch-Based Multi-label Classifier

Prantik Howlader, Srijan Das, Hieu Le, Dimitris Samaras

Abstract:

Incorporating pixel contextual information is critical for ac-
curate segmentation. In this paper, we show that an effective way to
incorporate contextual information is through a patch-based classifier.
This patch classifier is trained to identify classes present within an image
region, which facilitates the elimination of distractors and enhances the
classification of small object segments. Specifically, we introduce Multi-
scale Patch-based Multi-label Classifier (MPMC), a novel plug-in
module designed for existing semi-supervised segmentation (SSS) frame-
works. MPMC offers patch-level supervision, enabling the discrimina-
tion of pixel regions of different classes within a patch. Furthermore,
MPMC learns an adaptive pseudo-label weight, using patch-level classi-
fication to alleviate the impact of the teacher’s noisy pseudo-label super-
vision on the student. This lightweight module can be integrated into any
SSS framework, significantly enhancing their performance. We demon-
strate the efficacy of our proposed MPMC by integrating it into four
SSS methodologies and improving them across two natural image and
one medical segmentation dataset, notably improving the segmentation
results of the baselines across all the three datasets.