Naive-Student: Leveraging Semi-Supervised Learning in Video Sequences for Urban Scene Segmentation

By Liang-Chieh Chen, Raphael Gontijo Lopes, Bowen Cheng, Maxwell D. Collins, Ekin Dogus Cubuk, Barret Zoph, Hartwig Adam, Jonathon Shlens. 2020

121 citations recorded by the source corpus.

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

ZeroWaste Dataset: Towards Deformable Object Segmentation in Cluttered Scenes (2021, background): “Semi-supervised segmentation methods have been proposed to address such limitations by jointly learning from both annotated and unannotated images [11, 22, 29, 40, 42, 45].”
DetectoRS: Detecting Objects with Recursive Feature Pyramid and Switchable Atrous Convolution (2020, methodology): “As DetectoRS only detects things, we use the stuff predictions by DeepLabv3+ [15] with backbone WideResNet-41 [11, 76, 84].”
DenseCLIP: Extract Free Dense Labels from CLIP (2021, unknown): “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].”
k-means Mask Transformer (2022, methodology): “Therefore, we additionally perform a simple experiment to alleviate the over-ﬁtting issue by generating pseudo labels [12] on COCO unlabeled set.”
Pointly-Supervised Instance Segmentation (2021, background): “The gap between point and full mask supervision can be further decreased using selftraining paradigm [7, 47, 49, 56, 59, 62].”
MultiMAE: Multi-modal Multi-task Masked Autoencoders (2022, background): “These approaches have been applied to a variety of vision tasks such as image classification [65, 92, 95], object detection [104], and segmentation [15,104].”
Scaling Wide Residual Networks for Panoptic Segmentation (2020, methodology): “stage input size output size WR-41 [4] SWideRNet-(w1, w2, `) conv1 224×224 112×112 3×3, 64, stride 2 3×3, 64×w1, stride 2”
PseudoSeg: Designing Pseudo Labels for Semantic Segmentation (2020, background): “Recent exploration demonstrates improvement over high-data regime settings with large-scale data, including self-training (Chen et al., 2020a), backbone pre-training (Zhang et al., 2020a), and self-training (Zoph et al., 2020).”