Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313129917> ?p ?o ?g. }
- W4313129917 abstract "Multi-task indoor scene understanding is widely considered as an intriguing formulation, as the affinity of different tasks may lead to improved performance. In this paper, we tackle the new problem of Joint semantic, affordance and attribute parsing. However, successfully resolving it requires a model to capture long-range dependency, learn from weakly aligned data and properly balance sub-tasks during training. To this end, we propose an attention-based architecture named Cerberus and a tailored training framework. Our method effectively addresses aforementioned challenges and achieves state-of-the-art performance on all three tasks. Moreover, an in-depth analysis shows concept affinity consistent with human cognition, which inspires us to explore the possibility of weakly supervised learning. Surprisingly, Cerberus achieves strong results using only 0.1%–1% annotation. Visualizations further confirm that this success is credited to common attention maps across tasks. Code and models can be accessed at https://github.com/OPEN-AIR-SUN/Cerberus." @default.
- W4313129917 created "2023-01-06" @default.
- W4313129917 creator A5011913905 @default.
- W4313129917 creator A5025049450 @default.
- W4313129917 creator A5031438011 @default.
- W4313129917 creator A5038032242 @default.
- W4313129917 creator A5038258771 @default.
- W4313129917 date "2022-06-01" @default.
- W4313129917 modified "2023-09-30" @default.
- W4313129917 title "Cerberus Transformer: Joint Semantic, Affordance and Attribute Parsing" @default.
- W4313129917 cites W1905829557 @default.
- W4313129917 cites W2011511325 @default.
- W4313129917 cites W2054279472 @default.
- W4313129917 cites W2060727197 @default.
- W4313129917 cites W2068397385 @default.
- W4313129917 cites W2070148066 @default.
- W4313129917 cites W2082394523 @default.
- W4313129917 cites W2113107168 @default.
- W4313129917 cites W2294130536 @default.
- W4313129917 cites W2535516436 @default.
- W4313129917 cites W2536626143 @default.
- W4313129917 cites W2557465155 @default.
- W4313129917 cites W2560023338 @default.
- W4313129917 cites W2563705555 @default.
- W4313129917 cites W2741998915 @default.
- W4313129917 cites W2776622059 @default.
- W4313129917 cites W2962984928 @default.
- W4313129917 cites W2963591013 @default.
- W4313129917 cites W2965980545 @default.
- W4313129917 cites W2967870022 @default.
- W4313129917 cites W2971014764 @default.
- W4313129917 cites W3009205130 @default.
- W4313129917 cites W3010276570 @default.
- W4313129917 cites W3034225195 @default.
- W4313129917 cites W3034550906 @default.
- W4313129917 cites W3034868495 @default.
- W4313129917 cites W3035424742 @default.
- W4313129917 cites W3107485598 @default.
- W4313129917 cites W3108601100 @default.
- W4313129917 cites W3175844808 @default.
- W4313129917 cites W3198923361 @default.
- W4313129917 cites W4214520160 @default.
- W4313129917 doi "https://doi.org/10.1109/cvpr52688.2022.01903" @default.
- W4313129917 hasPublicationYear "2022" @default.
- W4313129917 type Work @default.
- W4313129917 citedByCount "5" @default.
- W4313129917 countsByYear W43131299172023 @default.
- W4313129917 crossrefType "proceedings-article" @default.
- W4313129917 hasAuthorship W4313129917A5011913905 @default.
- W4313129917 hasAuthorship W4313129917A5025049450 @default.
- W4313129917 hasAuthorship W4313129917A5031438011 @default.
- W4313129917 hasAuthorship W4313129917A5038032242 @default.
- W4313129917 hasAuthorship W4313129917A5038258771 @default.
- W4313129917 hasBestOaLocation W43131299172 @default.
- W4313129917 hasConcept C107457646 @default.
- W4313129917 hasConcept C119857082 @default.
- W4313129917 hasConcept C121332964 @default.
- W4313129917 hasConcept C127413603 @default.
- W4313129917 hasConcept C154945302 @default.
- W4313129917 hasConcept C162324750 @default.
- W4313129917 hasConcept C164883195 @default.
- W4313129917 hasConcept C165801399 @default.
- W4313129917 hasConcept C170154142 @default.
- W4313129917 hasConcept C18555067 @default.
- W4313129917 hasConcept C186644900 @default.
- W4313129917 hasConcept C187736073 @default.
- W4313129917 hasConcept C194995250 @default.
- W4313129917 hasConcept C204321447 @default.
- W4313129917 hasConcept C2776321320 @default.
- W4313129917 hasConcept C2780451532 @default.
- W4313129917 hasConcept C41008148 @default.
- W4313129917 hasConcept C62520636 @default.
- W4313129917 hasConcept C66322947 @default.
- W4313129917 hasConceptScore W4313129917C107457646 @default.
- W4313129917 hasConceptScore W4313129917C119857082 @default.
- W4313129917 hasConceptScore W4313129917C121332964 @default.
- W4313129917 hasConceptScore W4313129917C127413603 @default.
- W4313129917 hasConceptScore W4313129917C154945302 @default.
- W4313129917 hasConceptScore W4313129917C162324750 @default.
- W4313129917 hasConceptScore W4313129917C164883195 @default.
- W4313129917 hasConceptScore W4313129917C165801399 @default.
- W4313129917 hasConceptScore W4313129917C170154142 @default.
- W4313129917 hasConceptScore W4313129917C18555067 @default.
- W4313129917 hasConceptScore W4313129917C186644900 @default.
- W4313129917 hasConceptScore W4313129917C187736073 @default.
- W4313129917 hasConceptScore W4313129917C194995250 @default.
- W4313129917 hasConceptScore W4313129917C204321447 @default.
- W4313129917 hasConceptScore W4313129917C2776321320 @default.
- W4313129917 hasConceptScore W4313129917C2780451532 @default.
- W4313129917 hasConceptScore W4313129917C41008148 @default.
- W4313129917 hasConceptScore W4313129917C62520636 @default.
- W4313129917 hasConceptScore W4313129917C66322947 @default.
- W4313129917 hasLocation W43131299171 @default.
- W4313129917 hasLocation W43131299172 @default.
- W4313129917 hasOpenAccess W4313129917 @default.
- W4313129917 hasPrimaryLocation W43131299171 @default.
- W4313129917 hasRelatedWork W2020540721 @default.
- W4313129917 hasRelatedWork W2251128110 @default.
- W4313129917 hasRelatedWork W2251868338 @default.
- W4313129917 hasRelatedWork W2293457016 @default.