Matches in SemOpenAlex for { <https://semopenalex.org/work/W3171007011> ?p ?o ?g. }
Showing items 1 to 94 of
94
with 100 items per page.
- W3171007011 abstract "Siamese networks have become a common structure in various recent models for unsupervised visual representation learning. These models maximize the similarity between two augmentations of one image, subject to certain conditions for avoiding collapsing solutions. In this paper, we report surprising empirical results that simple Siamese networks can learn meaningful representations even using none of the following: (i) negative sample pairs, (ii) large batches, (iii) momentum encoders. Our experiments show that collapsing solutions do exist for the loss and structure, but a stop-gradient operation plays an essential role in preventing collapsing. We provide a hypothesis on the implication of stop-gradient, and further show proof-of-concept experiments verifying it. Our SimSiam method achieves competitive results on ImageNet and downstream tasks. We hope this simple baseline will motivate people to rethink the roles of Siamese architectures for unsupervised representation learning. Code is made available. <sup xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>1</sup>" @default.
- W3171007011 created "2021-06-22" @default.
- W3171007011 creator A5016095713 @default.
- W3171007011 creator A5049098969 @default.
- W3171007011 date "2021-06-01" @default.
- W3171007011 modified "2023-10-14" @default.
- W3171007011 title "Exploring Simple Siamese Representation Learning" @default.
- W3171007011 cites W2031489346 @default.
- W3171007011 cites W2108598243 @default.
- W3171007011 cites W2138621090 @default.
- W3171007011 cites W2145287260 @default.
- W3171007011 cites W2147800946 @default.
- W3171007011 cites W2194775991 @default.
- W3171007011 cites W2798991696 @default.
- W3171007011 cites W2963150697 @default.
- W3171007011 cites W2979579363 @default.
- W3171007011 cites W2987741655 @default.
- W3171007011 doi "https://doi.org/10.1109/cvpr46437.2021.01549" @default.
- W3171007011 hasPublicationYear "2021" @default.
- W3171007011 type Work @default.
- W3171007011 sameAs 3171007011 @default.
- W3171007011 citedByCount "1005" @default.
- W3171007011 countsByYear W31710070112020 @default.
- W3171007011 countsByYear W31710070112021 @default.
- W3171007011 countsByYear W31710070112022 @default.
- W3171007011 countsByYear W31710070112023 @default.
- W3171007011 crossrefType "proceedings-article" @default.
- W3171007011 hasAuthorship W3171007011A5016095713 @default.
- W3171007011 hasAuthorship W3171007011A5049098969 @default.
- W3171007011 hasBestOaLocation W31710070112 @default.
- W3171007011 hasConcept C103278499 @default.
- W3171007011 hasConcept C108583219 @default.
- W3171007011 hasConcept C111472728 @default.
- W3171007011 hasConcept C111919701 @default.
- W3171007011 hasConcept C115961682 @default.
- W3171007011 hasConcept C118505674 @default.
- W3171007011 hasConcept C119857082 @default.
- W3171007011 hasConcept C138885662 @default.
- W3171007011 hasConcept C154945302 @default.
- W3171007011 hasConcept C161191863 @default.
- W3171007011 hasConcept C177264268 @default.
- W3171007011 hasConcept C17744445 @default.
- W3171007011 hasConcept C199360897 @default.
- W3171007011 hasConcept C199539241 @default.
- W3171007011 hasConcept C204321447 @default.
- W3171007011 hasConcept C2776359362 @default.
- W3171007011 hasConcept C2776760102 @default.
- W3171007011 hasConcept C2777855551 @default.
- W3171007011 hasConcept C2780586882 @default.
- W3171007011 hasConcept C41008148 @default.
- W3171007011 hasConcept C59404180 @default.
- W3171007011 hasConcept C80444323 @default.
- W3171007011 hasConcept C94625758 @default.
- W3171007011 hasConceptScore W3171007011C103278499 @default.
- W3171007011 hasConceptScore W3171007011C108583219 @default.
- W3171007011 hasConceptScore W3171007011C111472728 @default.
- W3171007011 hasConceptScore W3171007011C111919701 @default.
- W3171007011 hasConceptScore W3171007011C115961682 @default.
- W3171007011 hasConceptScore W3171007011C118505674 @default.
- W3171007011 hasConceptScore W3171007011C119857082 @default.
- W3171007011 hasConceptScore W3171007011C138885662 @default.
- W3171007011 hasConceptScore W3171007011C154945302 @default.
- W3171007011 hasConceptScore W3171007011C161191863 @default.
- W3171007011 hasConceptScore W3171007011C177264268 @default.
- W3171007011 hasConceptScore W3171007011C17744445 @default.
- W3171007011 hasConceptScore W3171007011C199360897 @default.
- W3171007011 hasConceptScore W3171007011C199539241 @default.
- W3171007011 hasConceptScore W3171007011C204321447 @default.
- W3171007011 hasConceptScore W3171007011C2776359362 @default.
- W3171007011 hasConceptScore W3171007011C2776760102 @default.
- W3171007011 hasConceptScore W3171007011C2777855551 @default.
- W3171007011 hasConceptScore W3171007011C2780586882 @default.
- W3171007011 hasConceptScore W3171007011C41008148 @default.
- W3171007011 hasConceptScore W3171007011C59404180 @default.
- W3171007011 hasConceptScore W3171007011C80444323 @default.
- W3171007011 hasConceptScore W3171007011C94625758 @default.
- W3171007011 hasLocation W31710070111 @default.
- W3171007011 hasLocation W31710070112 @default.
- W3171007011 hasOpenAccess W3171007011 @default.
- W3171007011 hasPrimaryLocation W31710070111 @default.
- W3171007011 hasRelatedWork W2142795561 @default.
- W3171007011 hasRelatedWork W2379948177 @default.
- W3171007011 hasRelatedWork W2384362569 @default.
- W3171007011 hasRelatedWork W2611989081 @default.
- W3171007011 hasRelatedWork W2965925734 @default.
- W3171007011 hasRelatedWork W3048601286 @default.
- W3171007011 hasRelatedWork W4205302943 @default.
- W3171007011 hasRelatedWork W4375867731 @default.
- W3171007011 hasRelatedWork W2107949441 @default.
- W3171007011 hasRelatedWork W2181948922 @default.
- W3171007011 isParatext "false" @default.
- W3171007011 isRetracted "false" @default.
- W3171007011 magId "3171007011" @default.
- W3171007011 workType "article" @default.