More Like this

1. JGCL: Joint Self-Supervised and Supervised Graph Contrastive Learning

   https://doi.org/10.1145/3487553.3524722  

   Akkas, Selahattin ; Azad, Ariful ( April 2022 , WWW '22: Companion Proceedings of the Web Conference 2022)
2. Is self-supervised learning more robust than supervised learning?

   Zhong, Y. ; Tang, H. ; Chen, J. ; Peng, J. ; Wang, Y.-X. ( January 2022 , Proc ICML Workshop on Pre-training)
3. Self-Supervised Wasserstein Pseudo-Labeling for Semi-Supervised Image Classification

   https://doi.org/10.1109/CVPR46437.2021.01209  

   Taherkhani, Fariborz ; Dabouei, Ali ; Soleymani, Sobhan ; Dawson, Jeremy ; Nasrabadi, Nasser M. ( June 2021 , IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR))
4. Holistically-Attracted Wireframe Parsing: From Supervised to Self-Supervised Learning

   https://doi.org/10.1109/TPAMI.2023.3312749  

   Xue, Nan ; Wu, Tianfu ; Bai, Song ; Wang, Fu-Dong ; Xia, Gui-Song ; Zhang, Liangpei ; Torr, Philip H.S. ( January 2023 , IEEE Transactions on Pattern Analysis and Machine Intelligence)

   This article presents Holistically-Attracted Wireframe Parsing (HAWP), a method for geometric analysis of 2D images containing wireframes formed by line segments and junctions. HAWP utilizes a parsimonious Holistic Attraction (HAT) field representation that encodes line segments using a closed-form 4D geometric vector field. The proposed HAWP consists of three sequential components empowered by end-to-end and HAT-driven designs: (1) generating a dense set of line segments from HAT fields and endpoint proposals from heatmaps, (2) binding the dense line segments to sparse endpoint proposals to produce initial wireframes, and (3) filtering false positive proposals through a novel endpoint-decoupled line-of-interest aligning (EPD LOIAlign) module that captures the co-occurrence between endpoint proposals and HAT fields for better verification. Thanks to our novel designs, HAWPv2 shows strong performance in fully supervised learning, while HAWPv3 excels in self-supervised learning, achieving superior repeatability scores and efficient training (24 GPU hours on a single GPU). Furthermore, HAWPv3 exhibits a promising potential for wireframe parsing in out-of-distribution images without providing ground truth labels of wireframes. 

   more » « less
5. Mismatched Supervised Learning

   https://doi.org/10.1109/ICASSP43922.2022.9747362  

   Xian, Xun ; Hong, Mingyi ; Ding, Jie ( May 2022 , International Conference on Acoustics, Speech, & Signal Processing (ICASSP))