Matches in SemOpenAlex for { <https://semopenalex.org/work/W3082586907> ?p ?o ?g. }
- W3082586907 abstract "Deep learning-based scene text detection can achieve preferable performance, powered with sufficient labeled training data. However, manual labeling is time consuming and laborious. At the extreme, the corresponding annotated data are unavailable. Exploiting synthetic data is a very promising solution except for domain distribution mismatches between synthetic datasets and real datasets. To address the severe domain distribution mismatch, we propose a synthetic-to-real domain adaptation method for scene text detection, which transfers knowledge from synthetic data (source domain) to real data (target domain). In this paper, a text self-training (TST) method and adversarial text instance alignment (ATA) for domain adaptive scene text detection are introduced. ATA helps the network learn domain-invariant features by training a domain classifier in an adversarial manner. TST diminishes the adverse effects of false positives~(FPs) and false negatives~(FNs) from inaccurate pseudo-labels. Two components have positive effects on improving the performance of scene text detectors when adapting from synthetic-to-real scenes. We evaluate the proposed method by transferring from SynthText, VISD to ICDAR2015, ICDAR2013. The results demonstrate the effectiveness of the proposed method with up to 10% improvement, which has important exploration significance for domain adaptive scene text detection. Code is available at https://github.com/weijiawu/SyntoReal_STD" @default.
- W3082586907 created "2020-09-08" @default.
- W3082586907 creator A5034584642 @default.
- W3082586907 creator A5041031140 @default.
- W3082586907 creator A5081727769 @default.
- W3082586907 date "2020-09-03" @default.
- W3082586907 modified "2023-09-26" @default.
- W3082586907 title "Synthetic-to-Real Unsupervised Domain Adaptation for Scene Text Detection in the Wild" @default.
- W3082586907 cites W1488125194 @default.
- W3082586907 cites W1972065312 @default.
- W3082586907 cites W2008806374 @default.
- W3082586907 cites W2118045473 @default.
- W3082586907 cites W2142159465 @default.
- W3082586907 cites W2144554289 @default.
- W3082586907 cites W2149466042 @default.
- W3082586907 cites W2343052201 @default.
- W3082586907 cites W2519818067 @default.
- W3082586907 cites W2604735854 @default.
- W3082586907 cites W2605076167 @default.
- W3082586907 cites W2605982830 @default.
- W3082586907 cites W2785383245 @default.
- W3082586907 cites W2786559811 @default.
- W3082586907 cites W2792781829 @default.
- W3082586907 cites W2795124604 @default.
- W3082586907 cites W2795889831 @default.
- W3082586907 cites W2798593490 @default.
- W3082586907 cites W2798681837 @default.
- W3082586907 cites W2806327167 @default.
- W3082586907 cites W2810028092 @default.
- W3082586907 cites W2851091816 @default.
- W3082586907 cites W2873558679 @default.
- W3082586907 cites W2895281799 @default.
- W3082586907 cites W2902494497 @default.
- W3082586907 cites W2949788122 @default.
- W3082586907 cites W2953894958 @default.
- W3082586907 cites W2955889502 @default.
- W3082586907 cites W2962773189 @default.
- W3082586907 cites W2962804639 @default.
- W3082586907 cites W2962810613 @default.
- W3082586907 cites W2962935569 @default.
- W3082586907 cites W2963299454 @default.
- W3082586907 cites W2963647456 @default.
- W3082586907 cites W2963826681 @default.
- W3082586907 cites W2964115968 @default.
- W3082586907 cites W2964277612 @default.
- W3082586907 cites W2964294787 @default.
- W3082586907 cites W2966743431 @default.
- W3082586907 cites W2967591398 @default.
- W3082586907 cites W2983626510 @default.
- W3082586907 cites W2987563462 @default.
- W3082586907 cites W2990069979 @default.
- W3082586907 cites W2991626090 @default.
- W3082586907 cites W3013262970 @default.
- W3082586907 cites W3034907434 @default.
- W3082586907 doi "https://doi.org/10.48550/arxiv.2009.01766" @default.
- W3082586907 hasPublicationYear "2020" @default.
- W3082586907 type Work @default.
- W3082586907 sameAs 3082586907 @default.
- W3082586907 citedByCount "1" @default.
- W3082586907 countsByYear W30825869072021 @default.
- W3082586907 crossrefType "posted-content" @default.
- W3082586907 hasAuthorship W3082586907A5034584642 @default.
- W3082586907 hasAuthorship W3082586907A5041031140 @default.
- W3082586907 hasAuthorship W3082586907A5081727769 @default.
- W3082586907 hasBestOaLocation W30825869071 @default.
- W3082586907 hasConcept C112789634 @default.
- W3082586907 hasConcept C119857082 @default.
- W3082586907 hasConcept C134306372 @default.
- W3082586907 hasConcept C153180895 @default.
- W3082586907 hasConcept C154945302 @default.
- W3082586907 hasConcept C160920958 @default.
- W3082586907 hasConcept C2776145971 @default.
- W3082586907 hasConcept C2776434776 @default.
- W3082586907 hasConcept C2989486834 @default.
- W3082586907 hasConcept C33923547 @default.
- W3082586907 hasConcept C36503486 @default.
- W3082586907 hasConcept C37736160 @default.
- W3082586907 hasConcept C41008148 @default.
- W3082586907 hasConcept C51632099 @default.
- W3082586907 hasConcept C64869954 @default.
- W3082586907 hasConcept C95623464 @default.
- W3082586907 hasConceptScore W3082586907C112789634 @default.
- W3082586907 hasConceptScore W3082586907C119857082 @default.
- W3082586907 hasConceptScore W3082586907C134306372 @default.
- W3082586907 hasConceptScore W3082586907C153180895 @default.
- W3082586907 hasConceptScore W3082586907C154945302 @default.
- W3082586907 hasConceptScore W3082586907C160920958 @default.
- W3082586907 hasConceptScore W3082586907C2776145971 @default.
- W3082586907 hasConceptScore W3082586907C2776434776 @default.
- W3082586907 hasConceptScore W3082586907C2989486834 @default.
- W3082586907 hasConceptScore W3082586907C33923547 @default.
- W3082586907 hasConceptScore W3082586907C36503486 @default.
- W3082586907 hasConceptScore W3082586907C37736160 @default.
- W3082586907 hasConceptScore W3082586907C41008148 @default.
- W3082586907 hasConceptScore W3082586907C51632099 @default.
- W3082586907 hasConceptScore W3082586907C64869954 @default.
- W3082586907 hasConceptScore W3082586907C95623464 @default.
- W3082586907 hasLocation W30825869071 @default.
- W3082586907 hasOpenAccess W3082586907 @default.
- W3082586907 hasPrimaryLocation W30825869071 @default.