Matches in SemOpenAlex for { <https://semopenalex.org/work/W3003359010> ?p ?o ?g. }
- W3003359010 abstract "In this paper, we aim at automatically searching an efficient network architecture for dense image prediction. Particularly, we follow the encoder-decoder style and focus on designing a connectivity structure for the decoder. To achieve that, we design a densely connected network with learnable connections, named Fully Dense Network, which contains a large set of possible final connectivity structures. We then employ gradient descent to search the optimal connectivity from the dense connections. The search process is guided by a novel loss function, which pushes the weight of each connection to be binary and the connections to be sparse. The discovered connectivity achieves competitive results on two segmentation datasets, while runs more than three times faster and requires less than half parameters compared to the state-of-the-art methods. An extensive experiment shows that the discovered connectivity is compatible with various backbones and generalizes well to other dense image prediction tasks." @default.
- W3003359010 created "2020-02-07" @default.
- W3003359010 creator A5011772578 @default.
- W3003359010 creator A5028693655 @default.
- W3003359010 creator A5054504128 @default.
- W3003359010 date "2019-10-01" @default.
- W3003359010 modified "2023-10-09" @default.
- W3003359010 title "SparseMask: Differentiable Connectivity Learning for Dense Image Prediction" @default.
- W3003359010 cites W1903029394 @default.
- W3003359010 cites W1976047850 @default.
- W3003359010 cites W2037227137 @default.
- W3003359010 cites W2110158442 @default.
- W3003359010 cites W2124592697 @default.
- W3003359010 cites W2144794286 @default.
- W3003359010 cites W2161185676 @default.
- W3003359010 cites W2183341477 @default.
- W3003359010 cites W2194775991 @default.
- W3003359010 cites W2412782625 @default.
- W3003359010 cites W2560023338 @default.
- W3003359010 cites W2563705555 @default.
- W3003359010 cites W2732547613 @default.
- W3003359010 cites W2737258237 @default.
- W3003359010 cites W2799166040 @default.
- W3003359010 cites W2962860921 @default.
- W3003359010 cites W2963163009 @default.
- W3003359010 cites W2963299740 @default.
- W3003359010 cites W2963446712 @default.
- W3003359010 cites W2963727650 @default.
- W3003359010 cites W2963857746 @default.
- W3003359010 cites W2964081807 @default.
- W3003359010 cites W4255158661 @default.
- W3003359010 cites W845365781 @default.
- W3003359010 doi "https://doi.org/10.1109/iccv.2019.00687" @default.
- W3003359010 hasPublicationYear "2019" @default.
- W3003359010 type Work @default.
- W3003359010 sameAs 3003359010 @default.
- W3003359010 citedByCount "9" @default.
- W3003359010 countsByYear W30033590102020 @default.
- W3003359010 countsByYear W30033590102021 @default.
- W3003359010 countsByYear W30033590102023 @default.
- W3003359010 crossrefType "proceedings-article" @default.
- W3003359010 hasAuthorship W3003359010A5011772578 @default.
- W3003359010 hasAuthorship W3003359010A5028693655 @default.
- W3003359010 hasAuthorship W3003359010A5054504128 @default.
- W3003359010 hasBestOaLocation W30033590102 @default.
- W3003359010 hasConcept C111919701 @default.
- W3003359010 hasConcept C115961682 @default.
- W3003359010 hasConcept C118505674 @default.
- W3003359010 hasConcept C120665830 @default.
- W3003359010 hasConcept C121332964 @default.
- W3003359010 hasConcept C134306372 @default.
- W3003359010 hasConcept C153180895 @default.
- W3003359010 hasConcept C153258448 @default.
- W3003359010 hasConcept C154945302 @default.
- W3003359010 hasConcept C177264268 @default.
- W3003359010 hasConcept C192209626 @default.
- W3003359010 hasConcept C199360897 @default.
- W3003359010 hasConcept C202615002 @default.
- W3003359010 hasConcept C33923547 @default.
- W3003359010 hasConcept C41008148 @default.
- W3003359010 hasConcept C48372109 @default.
- W3003359010 hasConcept C50644808 @default.
- W3003359010 hasConcept C89600930 @default.
- W3003359010 hasConcept C94375191 @default.
- W3003359010 hasConcept C98045186 @default.
- W3003359010 hasConceptScore W3003359010C111919701 @default.
- W3003359010 hasConceptScore W3003359010C115961682 @default.
- W3003359010 hasConceptScore W3003359010C118505674 @default.
- W3003359010 hasConceptScore W3003359010C120665830 @default.
- W3003359010 hasConceptScore W3003359010C121332964 @default.
- W3003359010 hasConceptScore W3003359010C134306372 @default.
- W3003359010 hasConceptScore W3003359010C153180895 @default.
- W3003359010 hasConceptScore W3003359010C153258448 @default.
- W3003359010 hasConceptScore W3003359010C154945302 @default.
- W3003359010 hasConceptScore W3003359010C177264268 @default.
- W3003359010 hasConceptScore W3003359010C192209626 @default.
- W3003359010 hasConceptScore W3003359010C199360897 @default.
- W3003359010 hasConceptScore W3003359010C202615002 @default.
- W3003359010 hasConceptScore W3003359010C33923547 @default.
- W3003359010 hasConceptScore W3003359010C41008148 @default.
- W3003359010 hasConceptScore W3003359010C48372109 @default.
- W3003359010 hasConceptScore W3003359010C50644808 @default.
- W3003359010 hasConceptScore W3003359010C89600930 @default.
- W3003359010 hasConceptScore W3003359010C94375191 @default.
- W3003359010 hasConceptScore W3003359010C98045186 @default.
- W3003359010 hasLocation W30033590101 @default.
- W3003359010 hasLocation W30033590102 @default.
- W3003359010 hasOpenAccess W3003359010 @default.
- W3003359010 hasPrimaryLocation W30033590101 @default.
- W3003359010 hasRelatedWork W2033914206 @default.
- W3003359010 hasRelatedWork W2078528035 @default.
- W3003359010 hasRelatedWork W2275988210 @default.
- W3003359010 hasRelatedWork W2353011204 @default.
- W3003359010 hasRelatedWork W2358941527 @default.
- W3003359010 hasRelatedWork W2385621972 @default.
- W3003359010 hasRelatedWork W2394327295 @default.
- W3003359010 hasRelatedWork W3035587749 @default.
- W3003359010 hasRelatedWork W4220781934 @default.
- W3003359010 hasRelatedWork W4330337454 @default.
- W3003359010 isParatext "false" @default.