Matches in SemOpenAlex for { <https://semopenalex.org/work/W4320342844> ?p ?o ?g. }
Showing items 1 to 71 of
71
with 100 items per page.
- W4320342844 abstract "Self-paced curriculum learning (SCL) has demonstrated its great potential in computer vision, natural language processing, etc. During training, it implements easy-to-hard sampling based on online estimation of data difficulty. Most SCL methods commonly adopt a loss-based strategy of estimating data difficulty and deweighting the `hard' samples in the early training stage. While achieving success in a variety of applications, SCL stills confront two challenges in a medical image analysis task, such as universal lesion detection, featuring insufficient and highly class-imbalanced data: (i) the loss-based difficulty measurer is inaccurate; ii) the hard samples are under-utilized from a deweighting mechanism. To overcome these challenges, in this paper we propose a novel mixed-order self-paced curriculum learning (Mo-SCL) method. We integrate both uncertainty and loss to better estimate difficulty online and mix both hard and easy samples in the same mini-batch to appropriately alleviate the problem of under-utilization of hard samples. We provide a theoretical investigation of our method in the context of stochastic gradient descent optimization and extensive experiments based on the DeepLesion benchmark dataset for universal lesion detection (ULD). When applied to two state-of-the-art ULD methods, the proposed mixed-order SCL method can provide a free boost to lesion detection accuracy without extra special network designs." @default.
- W4320342844 created "2023-02-13" @default.
- W4320342844 creator A5001768407 @default.
- W4320342844 creator A5053949969 @default.
- W4320342844 creator A5084846204 @default.
- W4320342844 date "2023-02-09" @default.
- W4320342844 modified "2023-09-23" @default.
- W4320342844 title "Mixed-order self-paced curriculum learning for universal lesion detection" @default.
- W4320342844 doi "https://doi.org/10.48550/arxiv.2302.04677" @default.
- W4320342844 hasPublicationYear "2023" @default.
- W4320342844 type Work @default.
- W4320342844 citedByCount "0" @default.
- W4320342844 crossrefType "posted-content" @default.
- W4320342844 hasAuthorship W4320342844A5001768407 @default.
- W4320342844 hasAuthorship W4320342844A5053949969 @default.
- W4320342844 hasAuthorship W4320342844A5084846204 @default.
- W4320342844 hasBestOaLocation W43203428441 @default.
- W4320342844 hasConcept C119857082 @default.
- W4320342844 hasConcept C13280743 @default.
- W4320342844 hasConcept C136197465 @default.
- W4320342844 hasConcept C151730666 @default.
- W4320342844 hasConcept C153180895 @default.
- W4320342844 hasConcept C154945302 @default.
- W4320342844 hasConcept C15744967 @default.
- W4320342844 hasConcept C162324750 @default.
- W4320342844 hasConcept C185798385 @default.
- W4320342844 hasConcept C187736073 @default.
- W4320342844 hasConcept C19417346 @default.
- W4320342844 hasConcept C205649164 @default.
- W4320342844 hasConcept C206688291 @default.
- W4320342844 hasConcept C2779343474 @default.
- W4320342844 hasConcept C2780451532 @default.
- W4320342844 hasConcept C41008148 @default.
- W4320342844 hasConcept C47177190 @default.
- W4320342844 hasConcept C50644808 @default.
- W4320342844 hasConcept C86803240 @default.
- W4320342844 hasConceptScore W4320342844C119857082 @default.
- W4320342844 hasConceptScore W4320342844C13280743 @default.
- W4320342844 hasConceptScore W4320342844C136197465 @default.
- W4320342844 hasConceptScore W4320342844C151730666 @default.
- W4320342844 hasConceptScore W4320342844C153180895 @default.
- W4320342844 hasConceptScore W4320342844C154945302 @default.
- W4320342844 hasConceptScore W4320342844C15744967 @default.
- W4320342844 hasConceptScore W4320342844C162324750 @default.
- W4320342844 hasConceptScore W4320342844C185798385 @default.
- W4320342844 hasConceptScore W4320342844C187736073 @default.
- W4320342844 hasConceptScore W4320342844C19417346 @default.
- W4320342844 hasConceptScore W4320342844C205649164 @default.
- W4320342844 hasConceptScore W4320342844C206688291 @default.
- W4320342844 hasConceptScore W4320342844C2779343474 @default.
- W4320342844 hasConceptScore W4320342844C2780451532 @default.
- W4320342844 hasConceptScore W4320342844C41008148 @default.
- W4320342844 hasConceptScore W4320342844C47177190 @default.
- W4320342844 hasConceptScore W4320342844C50644808 @default.
- W4320342844 hasConceptScore W4320342844C86803240 @default.
- W4320342844 hasLocation W43203428441 @default.
- W4320342844 hasOpenAccess W4320342844 @default.
- W4320342844 hasPrimaryLocation W43203428441 @default.
- W4320342844 hasRelatedWork W1694201405 @default.
- W4320342844 hasRelatedWork W2787191226 @default.
- W4320342844 hasRelatedWork W2792987183 @default.
- W4320342844 hasRelatedWork W2983785000 @default.
- W4320342844 hasRelatedWork W2989932438 @default.
- W4320342844 hasRelatedWork W2990868490 @default.
- W4320342844 hasRelatedWork W3012093649 @default.
- W4320342844 hasRelatedWork W3213280251 @default.
- W4320342844 hasRelatedWork W4220975826 @default.
- W4320342844 hasRelatedWork W4317243171 @default.
- W4320342844 isParatext "false" @default.
- W4320342844 isRetracted "false" @default.
- W4320342844 workType "article" @default.