FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4285428718> ?p ?o ?g. }

• W4285428718 abstract "Recently, many semi-supervised object detection (SSOD) methods adopt teacher-student framework and have achieved state-of-the-art results. However, the teacher network is tightly coupled with the student network since the teacher is an exponential moving average (EMA) of the student, which causes a performance bottleneck. To address the coupling problem, we propose a Cycle Self-Training (CST) framework for SSOD, which consists of two teachers T1 and T2, two students S1 and S2. Based on these networks, a cycle self-training mechanism is built, i.e., S1$rightarrow $T1$rightarrow $S2$rightarrow $T2$rightarrow $S1. For S$rightarrow $T, we also utilize the EMA weights of the students to update the teachers. For T$rightarrow $S, instead of providing supervision for its own student S1(S2) directly, the teacher T1(T2) generates pseudo-labels for the student S2(S1), which looses the coupling effect. Moreover, owing to the property of EMA, the teacher is most likely to accumulate the biases from the student and make the mistakes irreversible. To mitigate the problem, we also propose a distribution consistency reweighting strategy, where pseudo-labels are reweighted based on distribution consistency across the teachers T1 and T2. With the strategy, the two students S2 and S1 can be trained robustly with noisy pseudo labels to avoid confirmation biases. Extensive experiments prove the superiority of CST by consistently improving the AP over the baseline and outperforming state-of-the-art methods by 2.1% absolute AP improvements with scarce labeled data." @default.
• W4285428718 created "2022-07-15" @default.
• W4285428718 creator A5011904390 @default.
• W4285428718 creator A5028926076 @default.
• W4285428718 creator A5048629905 @default.
• W4285428718 creator A5056168495 @default.
• W4285428718 creator A5067483347 @default.
• W4285428718 creator A5081953793 @default.
• W4285428718 date "2022-10-10" @default.
• W4285428718 modified "2023-09-30" @default.
• W4285428718 title "Cycle Self-Training for Semi-Supervised Object Detection with Distribution Consistency Reweighting" @default.
• W4285428718 cites W1861492603 @default.
• W4285428718 cites W2117539524 @default.
• W4285428718 cites W2194775991 @default.
• W4285428718 cites W2565639579 @default.
• W4285428718 cites W2895094948 @default.
• W4285428718 cites W2963351448 @default.
• W4285428718 cites W2963849369 @default.
• W4285428718 cites W2964241181 @default.
• W4285428718 cites W2982770724 @default.
• W4285428718 cites W2989604896 @default.
• W4285428718 cites W2990231018 @default.
• W4285428718 cites W3005650525 @default.
• W4285428718 cites W3011722050 @default.
• W4285428718 cites W3012573144 @default.
• W4285428718 cites W3034427487 @default.
• W4285428718 cites W3035160371 @default.
• W4285428718 cites W3035473155 @default.
• W4285428718 cites W3096609285 @default.
• W4285428718 cites W3127743092 @default.
• W4285428718 cites W3138516171 @default.
• W4285428718 cites W3158278463 @default.
• W4285428718 cites W3158661000 @default.
• W4285428718 cites W3165924303 @default.
• W4285428718 cites W3172507542 @default.
• W4285428718 cites W3173770676 @default.
• W4285428718 cites W3176748778 @default.
• W4285428718 cites W3178291178 @default.
• W4285428718 cites W4288083516 @default.
• W4285428718 cites W4312377253 @default.
• W4285428718 cites W639708223 @default.
• W4285428718 doi "https://doi.org/10.1145/3503161.3548040" @default.
• W4285428718 hasPublicationYear "2022" @default.
• W4285428718 type Work @default.
• W4285428718 citedByCount "1" @default.
• W4285428718 countsByYear W42854287182023 @default.
• W4285428718 crossrefType "proceedings-article" @default.
• W4285428718 hasAuthorship W4285428718A5011904390 @default.
• W4285428718 hasAuthorship W4285428718A5028926076 @default.
• W4285428718 hasAuthorship W4285428718A5048629905 @default.
• W4285428718 hasAuthorship W4285428718A5056168495 @default.
• W4285428718 hasAuthorship W4285428718A5067483347 @default.
• W4285428718 hasAuthorship W4285428718A5081953793 @default.
• W4285428718 hasBestOaLocation W42854287181 @default.
• W4285428718 hasConcept C110121322 @default.
• W4285428718 hasConcept C111368507 @default.
• W4285428718 hasConcept C12725497 @default.
• W4285428718 hasConcept C127313418 @default.
• W4285428718 hasConcept C127413603 @default.
• W4285428718 hasConcept C131584629 @default.
• W4285428718 hasConcept C134306372 @default.
• W4285428718 hasConcept C149635348 @default.
• W4285428718 hasConcept C154945302 @default.
• W4285428718 hasConcept C2776436953 @default.
• W4285428718 hasConcept C2780513914 @default.
• W4285428718 hasConcept C2781238097 @default.
• W4285428718 hasConcept C33923547 @default.
• W4285428718 hasConcept C41008148 @default.
• W4285428718 hasConcept C78519656 @default.
• W4285428718 hasConceptScore W4285428718C110121322 @default.
• W4285428718 hasConceptScore W4285428718C111368507 @default.
• W4285428718 hasConceptScore W4285428718C12725497 @default.
• W4285428718 hasConceptScore W4285428718C127313418 @default.
• W4285428718 hasConceptScore W4285428718C127413603 @default.
• W4285428718 hasConceptScore W4285428718C131584629 @default.
• W4285428718 hasConceptScore W4285428718C134306372 @default.
• W4285428718 hasConceptScore W4285428718C149635348 @default.
• W4285428718 hasConceptScore W4285428718C154945302 @default.
• W4285428718 hasConceptScore W4285428718C2776436953 @default.
• W4285428718 hasConceptScore W4285428718C2780513914 @default.
• W4285428718 hasConceptScore W4285428718C2781238097 @default.
• W4285428718 hasConceptScore W4285428718C33923547 @default.
• W4285428718 hasConceptScore W4285428718C41008148 @default.
• W4285428718 hasConceptScore W4285428718C78519656 @default.
• W4285428718 hasFunder F4320321001 @default.
• W4285428718 hasLocation W42854287181 @default.
• W4285428718 hasLocation W42854287182 @default.
• W4285428718 hasOpenAccess W4285428718 @default.
• W4285428718 hasPrimaryLocation W42854287181 @default.
• W4285428718 hasRelatedWork W1603231086 @default.
• W4285428718 hasRelatedWork W2087937280 @default.
• W4285428718 hasRelatedWork W2353647904 @default.
• W4285428718 hasRelatedWork W2354251581 @default.
• W4285428718 hasRelatedWork W2357461155 @default.
• W4285428718 hasRelatedWork W2361361118 @default.
• W4285428718 hasRelatedWork W2384129116 @default.
• W4285428718 hasRelatedWork W2607965789 @default.
• W4285428718 hasRelatedWork W3145924829 @default.
• W4285428718 hasRelatedWork W3152267458 @default.
• W4285428718 isParatext "false" @default.

• next