GuidedMix-Net: Learning to Improve Pseudo Masks Using Labeled Images as Reference

29 Jun 2021  ·  Peng Tu, Yawen Huang, Rongrong Ji, Feng Zheng, Ling Shao ·

Semi-supervised learning is a challenging problem which aims to construct a model by learning from a limited number of labeled examples. Numerous methods have been proposed to tackle this problem, with most focusing on utilizing the predictions of unlabeled instances consistency alone to regularize networks... However, treating labeled and unlabeled data separately often leads to the discarding of mass prior knowledge learned from the labeled examples, and failure to mine the feature interaction between the labeled and unlabeled image pairs. In this paper, we propose a novel method for semi-supervised semantic segmentation named GuidedMix-Net, by leveraging labeled information to guide the learning of unlabeled instances. Specifically, we first introduce a feature alignment objective between labeled and unlabeled data to capture potentially similar image pairs and then generate mixed inputs from them. The proposed mutual information transfer (MITrans), based on the cluster assumption, is shown to be a powerful knowledge module for further progressive refining features of unlabeled data in the mixed data space. To take advantage of the labeled examples and guide unlabeled data learning, we further propose a mask generation module to generate high-quality pseudo masks for the unlabeled data. Along with supervised learning for labeled data, the prediction of unlabeled data is jointly learned with the generated pseudo masks from the mixed data. Extensive experiments on PASCAL VOC 2012, PASCAL-Context and Cityscapes demonstrate the effectiveness of our GuidedMix-Net, which achieves competitive segmentation accuracy and significantly improves the mIoU by +7$\%$ compared to previous state-of-the-art approaches. read more

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Semi-Supervised Semantic Segmentation Cityscapes 100 samples labeled GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) Validation mIoU 56.9% # 5
Semi-Supervised Semantic Segmentation Cityscapes 12.5% labeled GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) Validation mIoU 65.8% # 4
Semi-Supervised Semantic Segmentation Cityscapes 25% labeled GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) Validation mIoU 67.5% # 5
Semi-Supervised Semantic Segmentation Cityscapes 50% labeled GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) Validation mIoU 69.8% # 3
Semi-Supervised Semantic Segmentation PASCAL Context 12.5% labeled GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) Validation mIoU 40.3% # 1
Semi-Supervised Semantic Segmentation PASCAL Context 25% labeled GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) Validation mIoU 41.7% # 1
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 1000 labels GuidedMix-Net(DeepLab v2 with ResNet50, ImageNet pretrained) Validation mIoU 68.1% # 1
Semi-Supervised Semantic Segmentation Pascal VOC 2012 12.5% labeled GuidedMix-Net(DeepLabv2 with ResNet101, ImageNet pretrained) Validation mIoU 73.4% # 4
Semi-Supervised Semantic Segmentation Pascal VOC 2012 12.5% labeled GuidedMix-Net(DeepLab v2 with ResNet101, input size: 512x512 with multi-scale and flip, image-net pretrained ImageNet pretrained) Validation mIoU 76.4% # 2
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 1464 labels GuidedMix-Net(DeepLab v2 with ResNet50, ImageNet pretrained) Validation mIoU 73.7 # 2
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 25% labeled GuidedMix-Net(DeepLab v2 with ResNet101, input-size: 512x512 with multi-scale and flip, ImageNet pretrained) Validation mIoU 77.8% # 1
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 25% labeled GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) Validation mIoU 75.5% # 3
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 50% GuidedMix-Net(DeepLab v2 with ResNet101, input-size: 512x512 with multi-scale and flip, ImageNet pretrained) Validation mIoU 78.2% # 1
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 50% GuidedMix-Net(DeepLab v2 with ResNet101, ImageNet pretrained) Validation mIoU 76.5% # 2
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 500 labels GuidedMix-Net(DeepLab v2 with ResNet50, ImageNet pretrained) Validation mIoU 65.4% # 1

Methods


No methods listed for this paper. Add relevant methods here