Matches in SemOpenAlex for { <https://semopenalex.org/work/W3211791819> ?p ?o ?g. }
Showing items 1 to 67 of
67
with 100 items per page.
- W3211791819 abstract "Due to the limited and even imbalanced data, semi-supervised semantic segmentation tends to have poor performance on some certain categories, e.g., tailed categories in Cityscapes dataset which exhibits a long-tailed label distribution. Existing approaches almost all neglect this problem, and treat categories equally. Some popular approaches such as consistency regularization or pseudo-labeling may even harm the learning of under-performing categories, that the predictions or pseudo labels of these categories could be too inaccurate to guide the learning on the unlabeled data. In this paper, we look into this problem, and propose a novel framework for semi-supervised semantic segmentation, named adaptive equalization learning (AEL). AEL adaptively balances the training of well and badly performed categories, with a confidence bank to dynamically track category-wise performance during training. The confidence bank is leveraged as an indicator to tilt training towards under-performing categories, instantiated in three strategies: 1) adaptive Copy-Paste and CutMix data augmentation approaches which give more chance for under-performing categories to be copied or cut; 2) an adaptive data sampling approach to encourage pixels from under-performing category to be sampled; 3) a simple yet effective re-weighting method to alleviate the training noise raised by pseudo-labeling. Experimentally, AEL outperforms the state-of-the-art methods by a large margin on the Cityscapes and Pascal VOC benchmarks under various data partition protocols. Code is available at https://github.com/hzhupku/SemiSeg-AEL" @default.
- W3211791819 created "2021-11-22" @default.
- W3211791819 creator A5004122403 @default.
- W3211791819 creator A5055723755 @default.
- W3211791819 creator A5068464271 @default.
- W3211791819 creator A5068656698 @default.
- W3211791819 creator A5090180284 @default.
- W3211791819 creator A5090973869 @default.
- W3211791819 date "2021-10-11" @default.
- W3211791819 modified "2023-09-26" @default.
- W3211791819 title "Semi-Supervised Semantic Segmentation via Adaptive Equalization Learning" @default.
- W3211791819 doi "https://doi.org/10.48550/arxiv.2110.05474" @default.
- W3211791819 hasPublicationYear "2021" @default.
- W3211791819 type Work @default.
- W3211791819 sameAs 3211791819 @default.
- W3211791819 citedByCount "0" @default.
- W3211791819 crossrefType "posted-content" @default.
- W3211791819 hasAuthorship W3211791819A5004122403 @default.
- W3211791819 hasAuthorship W3211791819A5055723755 @default.
- W3211791819 hasAuthorship W3211791819A5068464271 @default.
- W3211791819 hasAuthorship W3211791819A5068656698 @default.
- W3211791819 hasAuthorship W3211791819A5090180284 @default.
- W3211791819 hasAuthorship W3211791819A5090973869 @default.
- W3211791819 hasBestOaLocation W32117918191 @default.
- W3211791819 hasConcept C119857082 @default.
- W3211791819 hasConcept C126838900 @default.
- W3211791819 hasConcept C154945302 @default.
- W3211791819 hasConcept C183115368 @default.
- W3211791819 hasConcept C199360897 @default.
- W3211791819 hasConcept C2776145971 @default.
- W3211791819 hasConcept C2776436953 @default.
- W3211791819 hasConcept C41008148 @default.
- W3211791819 hasConcept C58973888 @default.
- W3211791819 hasConcept C71924100 @default.
- W3211791819 hasConcept C75608658 @default.
- W3211791819 hasConcept C89600930 @default.
- W3211791819 hasConcept C95623464 @default.
- W3211791819 hasConceptScore W3211791819C119857082 @default.
- W3211791819 hasConceptScore W3211791819C126838900 @default.
- W3211791819 hasConceptScore W3211791819C154945302 @default.
- W3211791819 hasConceptScore W3211791819C183115368 @default.
- W3211791819 hasConceptScore W3211791819C199360897 @default.
- W3211791819 hasConceptScore W3211791819C2776145971 @default.
- W3211791819 hasConceptScore W3211791819C2776436953 @default.
- W3211791819 hasConceptScore W3211791819C41008148 @default.
- W3211791819 hasConceptScore W3211791819C58973888 @default.
- W3211791819 hasConceptScore W3211791819C71924100 @default.
- W3211791819 hasConceptScore W3211791819C75608658 @default.
- W3211791819 hasConceptScore W3211791819C89600930 @default.
- W3211791819 hasConceptScore W3211791819C95623464 @default.
- W3211791819 hasLocation W32117918191 @default.
- W3211791819 hasOpenAccess W3211791819 @default.
- W3211791819 hasPrimaryLocation W32117918191 @default.
- W3211791819 hasRelatedWork W2107003417 @default.
- W3211791819 hasRelatedWork W2156671090 @default.
- W3211791819 hasRelatedWork W2160952319 @default.
- W3211791819 hasRelatedWork W2538661024 @default.
- W3211791819 hasRelatedWork W2979433110 @default.
- W3211791819 hasRelatedWork W2981212515 @default.
- W3211791819 hasRelatedWork W3077461311 @default.
- W3211791819 hasRelatedWork W3181161034 @default.
- W3211791819 hasRelatedWork W4294974824 @default.
- W3211791819 hasRelatedWork W4313527211 @default.
- W3211791819 isParatext "false" @default.
- W3211791819 isRetracted "false" @default.
- W3211791819 magId "3211791819" @default.
- W3211791819 workType "article" @default.