Semi-Supervised Semantic Segmentation with Cross Pseudo Supervision

By Xiaokang Chen, Yuhui Yuan, Gang Zeng, Jingdong Wang. 2021

1,179 citations recorded by the source corpus.

In this paper, we study the semi-supervised semantic segmentation problem via exploring both labeled data and extra unlabeled data. We propose a novel consistency regularization approach, called cross pseudo supervision (CPS). Our approach imposes the consistency on two segmentation networks perturbed with different initialization for the same input image. The pseudo one-hot label map, output from one perturbed segmentation network, is used to supervise the other segmentation network with the standard cross-entropy loss, and vice versa. The CPS consistency has two roles: encourage high similarity between the predictions of two perturbed networks for the same input image, and expand training data by using the unlabeled data with pseudo labels. Experiment results show that our approach achieves the state-of-the-art semi-supervised segmentation performance on Cityscapes and PASCAL VOC 2012.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Large Language Models Can Self-Improve (2022, neutral): “The key idea is to assign pseudo labels from a learned classifier to unlabeled data, and use these pseudo-labeled examples to further improve the original model training, e.g., (RoyChowdhury et al., 2019; Xie et al., 2020; He et al., 2020; Chen et al., 2021).”
Novel Class Discovery in Semantic Segmentation (2021, methodology): “Several settings are proposed to alleviate the cost, such as semi-supervised learning (SSL), unsupervised learning (USL), etc .”
Semi-Supervised Medical Image Segmentation via Cross Teaching between CNN and Transformer (2021, methodology): “A lot of semi-supervised methods have been proposed for medical image analysis, including pseudo-labelling (Wang et al., 2021a; Bai et al., 2017; Chen et al., 2021b), deep co-training (Qiao et al., 2018; Zhou et al., 2019), deep adversarial learning (Zhang et al., 2017; Hu et al., 2020), few-shot…”
Self-Training: A Survey (2022, positive): “In the wake of works utilizing only a single classifier in self-training algorithms, new studies have been proposed with the use of two classifiers, where each model learns on the output of the other [Xie et al., 2020b; Chen et al., 2021; Karamanolakis et al., 2021].”
Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation (2022, methodology): “We follow two techniques in previous SOTA works [13,28,41,61,65], i.”
Mutual consistency learning for semi-supervised medical image segmentation (2021, methodology): “Furthermore, the pseudo label learning (Lee et al., 2013; Chen et al., 2021) often employs a sharpening function or a ﬁxed threshold to convert probability maps into pseudo labels.”
Semi-Supervised Semantic Segmentation Using Unreliable Pseudo-Labels (2022, methodology): “To alleviate this problem, many attempts [1,4,11,17,23,34,44,49] have been made towards semi-supervised semantic segmentation, which learns a model with only a few labeled samples and numerous unlabeled ones.”
SimCVD: Simple Contrastive Voxel-Wise Representation Distillation for Semi-Supervised Medical Image Segmentation (2021, result): “We compare SimCVD with published results from previous state-of-the-art semi-supervised segmentation methods, including V-Net [7], MT [31], DAN [15], CPS [60], Entropy Mini [61], UA-MT [20], ICT [62], SASSNet [21], DCT [23], and Chaitanya et al.[24] on the LA dataset in two labeled ratio settings (i”
Extract Free Dense Labels from CLIP (2021, methodology): “Self-training has firstly become popular in the semisupervised classification task [33,29,35,43] and is also recently applied in the semi-supervised/zero-shot semantic segmentation settings [27,37,4,28,40,36,56,8].”
Semi-Supervised and Unsupervised Deep Visual Learning: A Survey (2022, methodology): “, semantic segmentation [32], [202], object detection [30], [203], unsupervised domain adaptation [204], [205], pose estimation [34], [206], 3D scene understanding [207], video recognition [150], [208], etc.”
ST++: Make Self-trainingWork Better for Semi-supervised Semantic Segmentation (2021, methodology): “Therefore, also inspired by the success in SSL, subsequent methods [1, 12, 18, 24, 25, 31, 41, 55, 59, 61, 62, 66] manage to tackle this task with simpler mechanisms, such as enforcing similar predictions under multiple perturbed embeddings [41], under two different contextual crops [31], and betwee”
DenseCLIP: Extract Free Dense Labels from CLIP (2021, neutral): “Self-training has ﬁrstly become popular in the semi-supervised classiﬁcation task [29, 33, 35, 43] and is also recently applied in the semi-supervised/zero-shot semantic segmentation settings [4, 10, 27, 28, 36, 37, 40, 56].”
ZegCLIP: Towards Adapting CLIP for Zero-shot Semantic Segmentation (2022, background): “Semantic segmentation is one of the fundamental tasks in the computer vision field, which aims to predict the category of each pixel of an image [7, 15, 31, 41].”