Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386075981> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W4386075981 abstract "In this work, we revisit the weak-to-strong consistency framework, popularized by FixMatch from semi-supervised classification, where the prediction of a weakly perturbed image serves as supervision for its strongly perturbed version. Intriguingly, we observe that such a simple pipeline already achieves competitive results against recent advanced works, when transferred to our segmentation scenario. Its success heavily relies on the manual design of strong data augmentations, however, which may be limited and inadequate to explore a broader perturbation space. Motivated by this, we propose an auxiliary feature perturbation stream as a supplement, leading to an expanded perturbation space. On the other, to sufficiently probe original image-level augmentations, we present a dual-stream perturbation technique, enabling two strong views to be simultaneously guided by a common weak view. Consequently, our overall Unified Dual-Stream Perturbations approach (UniMatch) surpasses all existing methods significantly across all evaluation protocols on the Pascal, Cityscapes, and COCO benchmarks. Its superiority is also demonstrated in remote sensing interpretation and medical image analysis. We hope our reproduced FixMatch and our results can inspire more future works." @default.
- W4386075981 created "2023-08-23" @default.
- W4386075981 creator A5038009961 @default.
- W4386075981 creator A5039785210 @default.
- W4386075981 creator A5047297176 @default.
- W4386075981 creator A5055917015 @default.
- W4386075981 creator A5057842914 @default.
- W4386075981 date "2023-06-01" @default.
- W4386075981 modified "2023-10-11" @default.
- W4386075981 title "Revisiting Weak-to-Strong Consistency in Semi-Supervised Semantic Segmentation" @default.
- W4386075981 cites W1903029394 @default.
- W4386075981 cites W2037227137 @default.
- W4386075981 cites W2048679005 @default.
- W4386075981 cites W2079057609 @default.
- W4386075981 cites W2144794286 @default.
- W4386075981 cites W2194775991 @default.
- W4386075981 cites W2340897893 @default.
- W4386075981 cites W2560023338 @default.
- W4386075981 cites W2620998106 @default.
- W4386075981 cites W2804047627 @default.
- W4386075981 cites W2963727650 @default.
- W4386075981 cites W3035680157 @default.
- W4386075981 cites W3091002423 @default.
- W4386075981 cites W3109620645 @default.
- W4386075981 cites W3113960880 @default.
- W4386075981 cites W3157653192 @default.
- W4386075981 cites W3167306551 @default.
- W4386075981 cites W3171007011 @default.
- W4386075981 cites W3171581326 @default.
- W4386075981 cites W3172507542 @default.
- W4386075981 cites W3172949220 @default.
- W4386075981 cites W3180500590 @default.
- W4386075981 cites W3180562345 @default.
- W4386075981 cites W3195975500 @default.
- W4386075981 cites W3204140421 @default.
- W4386075981 cites W4214567023 @default.
- W4386075981 cites W4214899942 @default.
- W4386075981 cites W4221161877 @default.
- W4386075981 cites W4285602983 @default.
- W4386075981 cites W4312310512 @default.
- W4386075981 cites W4312328737 @default.
- W4386075981 cites W4312651959 @default.
- W4386075981 doi "https://doi.org/10.1109/cvpr52729.2023.00699" @default.
- W4386075981 hasPublicationYear "2023" @default.
- W4386075981 type Work @default.
- W4386075981 citedByCount "4" @default.
- W4386075981 countsByYear W43860759812023 @default.
- W4386075981 crossrefType "proceedings-article" @default.
- W4386075981 hasAuthorship W4386075981A5038009961 @default.
- W4386075981 hasAuthorship W4386075981A5039785210 @default.
- W4386075981 hasAuthorship W4386075981A5047297176 @default.
- W4386075981 hasAuthorship W4386075981A5055917015 @default.
- W4386075981 hasAuthorship W4386075981A5057842914 @default.
- W4386075981 hasConcept C11413529 @default.
- W4386075981 hasConcept C119857082 @default.
- W4386075981 hasConcept C121332964 @default.
- W4386075981 hasConcept C154945302 @default.
- W4386075981 hasConcept C177918212 @default.
- W4386075981 hasConcept C199360897 @default.
- W4386075981 hasConcept C2776436953 @default.
- W4386075981 hasConcept C41008148 @default.
- W4386075981 hasConcept C62520636 @default.
- W4386075981 hasConcept C75608658 @default.
- W4386075981 hasConcept C80444323 @default.
- W4386075981 hasConcept C89600930 @default.
- W4386075981 hasConceptScore W4386075981C11413529 @default.
- W4386075981 hasConceptScore W4386075981C119857082 @default.
- W4386075981 hasConceptScore W4386075981C121332964 @default.
- W4386075981 hasConceptScore W4386075981C154945302 @default.
- W4386075981 hasConceptScore W4386075981C177918212 @default.
- W4386075981 hasConceptScore W4386075981C199360897 @default.
- W4386075981 hasConceptScore W4386075981C2776436953 @default.
- W4386075981 hasConceptScore W4386075981C41008148 @default.
- W4386075981 hasConceptScore W4386075981C62520636 @default.
- W4386075981 hasConceptScore W4386075981C75608658 @default.
- W4386075981 hasConceptScore W4386075981C80444323 @default.
- W4386075981 hasConceptScore W4386075981C89600930 @default.
- W4386075981 hasFunder F4320321001 @default.
- W4386075981 hasLocation W43860759811 @default.
- W4386075981 hasOpenAccess W4386075981 @default.
- W4386075981 hasPrimaryLocation W43860759811 @default.
- W4386075981 hasRelatedWork W2066229105 @default.
- W4386075981 hasRelatedWork W2140468882 @default.
- W4386075981 hasRelatedWork W2961085424 @default.
- W4386075981 hasRelatedWork W3042343907 @default.
- W4386075981 hasRelatedWork W3209312100 @default.
- W4386075981 hasRelatedWork W4231350079 @default.
- W4386075981 hasRelatedWork W4244612519 @default.
- W4386075981 hasRelatedWork W4286681602 @default.
- W4386075981 hasRelatedWork W4306674287 @default.
- W4386075981 hasRelatedWork W4224009465 @default.
- W4386075981 isParatext "false" @default.
- W4386075981 isRetracted "false" @default.
- W4386075981 workType "article" @default.