Matches in SemOpenAlex for { <https://semopenalex.org/work/W3140819097> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W3140819097 abstract "Labeled data is a critical resource for training and evaluating machine learning models. However, many real-life datasets are only partially labeled. We propose a semi-supervised machine learning training strategy to improve event detection performance on sequential data, such as video recordings, when only sparse labels are available, such as event start times without their corresponding end times. Our method uses noisy guesses of the events’ end times to train event detection models. Depending on how conservative these guesses are, mislabeled samples may be introduced into the training set. We further propose a mathematical model for explaining and estimating the evolution of the classification performance for increasingly noisier end time estimates. We show that neural networks can improve their detection performance by leveraging more training data with less conservative approximations despite the higher proportion of incorrect labels. We adapt sequential versions of CIFAR-10 and MNIST, and use the Berkeley MHAD and HMBD51 video datasets to empirically evaluate our method, and find that our risk-tolerant strategy outperforms conservative estimates by 3.5 points of mean average precision for CIFAR, 30 points for MNIST, 3 points for MHAD, and 14 points for HMBD51. Then, we leverage the proposed training strategy to tackle a real-life application: processing continuous video recordings of epilepsy patients, and show that our method outperforms baseline labeling methods by 17 points of average precision, and reaches a classification performance similar to that of fully supervised models. We share part of the code for this article at the following repository: fpgdubost/CIFAR-10-Sparsely-Labeled-Sequential-Data." @default.
- W3140819097 created "2021-04-13" @default.
- W3140819097 creator A5004336107 @default.
- W3140819097 creator A5023339664 @default.
- W3140819097 creator A5044309572 @default.
- W3140819097 creator A5064184284 @default.
- W3140819097 creator A5077447259 @default.
- W3140819097 creator A5082175474 @default.
- W3140819097 date "2023-01-01" @default.
- W3140819097 modified "2023-09-26" @default.
- W3140819097 title "Semi-Supervised Learning for Sparsely-Labeled Sequential Data: Application to Healthcare Video Processing" @default.
- W3140819097 cites W2010215165 @default.
- W3140819097 cites W2031033421 @default.
- W3140819097 cites W2082704080 @default.
- W3140819097 cites W2108598243 @default.
- W3140819097 cites W2126579184 @default.
- W3140819097 cites W2132886902 @default.
- W3140819097 cites W2168571645 @default.
- W3140819097 cites W2340897893 @default.
- W3140819097 cites W2746791238 @default.
- W3140819097 cites W2963155035 @default.
- W3140819097 cites W2979433110 @default.
- W3140819097 cites W3000043583 @default.
- W3140819097 cites W3008257486 @default.
- W3140819097 cites W3016911444 @default.
- W3140819097 cites W3035524453 @default.
- W3140819097 cites W3036586801 @default.
- W3140819097 cites W3100570787 @default.
- W3140819097 cites W3101604456 @default.
- W3140819097 cites W3118508703 @default.
- W3140819097 doi "https://doi.org/10.1109/wacv56688.2023.00193" @default.
- W3140819097 hasPublicationYear "2023" @default.
- W3140819097 type Work @default.
- W3140819097 sameAs 3140819097 @default.
- W3140819097 citedByCount "0" @default.
- W3140819097 crossrefType "proceedings-article" @default.
- W3140819097 hasAuthorship W3140819097A5004336107 @default.
- W3140819097 hasAuthorship W3140819097A5023339664 @default.
- W3140819097 hasAuthorship W3140819097A5044309572 @default.
- W3140819097 hasAuthorship W3140819097A5064184284 @default.
- W3140819097 hasAuthorship W3140819097A5077447259 @default.
- W3140819097 hasAuthorship W3140819097A5082175474 @default.
- W3140819097 hasBestOaLocation W31408190972 @default.
- W3140819097 hasConcept C119857082 @default.
- W3140819097 hasConcept C121332964 @default.
- W3140819097 hasConcept C124101348 @default.
- W3140819097 hasConcept C153083717 @default.
- W3140819097 hasConcept C153180895 @default.
- W3140819097 hasConcept C154945302 @default.
- W3140819097 hasConcept C177264268 @default.
- W3140819097 hasConcept C190502265 @default.
- W3140819097 hasConcept C199360897 @default.
- W3140819097 hasConcept C21080849 @default.
- W3140819097 hasConcept C2776145971 @default.
- W3140819097 hasConcept C2779662365 @default.
- W3140819097 hasConcept C41008148 @default.
- W3140819097 hasConcept C50644808 @default.
- W3140819097 hasConcept C62520636 @default.
- W3140819097 hasConcept C67186912 @default.
- W3140819097 hasConcept C77088390 @default.
- W3140819097 hasConceptScore W3140819097C119857082 @default.
- W3140819097 hasConceptScore W3140819097C121332964 @default.
- W3140819097 hasConceptScore W3140819097C124101348 @default.
- W3140819097 hasConceptScore W3140819097C153083717 @default.
- W3140819097 hasConceptScore W3140819097C153180895 @default.
- W3140819097 hasConceptScore W3140819097C154945302 @default.
- W3140819097 hasConceptScore W3140819097C177264268 @default.
- W3140819097 hasConceptScore W3140819097C190502265 @default.
- W3140819097 hasConceptScore W3140819097C199360897 @default.
- W3140819097 hasConceptScore W3140819097C21080849 @default.
- W3140819097 hasConceptScore W3140819097C2776145971 @default.
- W3140819097 hasConceptScore W3140819097C2779662365 @default.
- W3140819097 hasConceptScore W3140819097C41008148 @default.
- W3140819097 hasConceptScore W3140819097C50644808 @default.
- W3140819097 hasConceptScore W3140819097C62520636 @default.
- W3140819097 hasConceptScore W3140819097C67186912 @default.
- W3140819097 hasConceptScore W3140819097C77088390 @default.
- W3140819097 hasFunder F4320335322 @default.
- W3140819097 hasLocation W31408190971 @default.
- W3140819097 hasLocation W31408190972 @default.
- W3140819097 hasOpenAccess W3140819097 @default.
- W3140819097 hasPrimaryLocation W31408190971 @default.
- W3140819097 hasRelatedWork W2292254049 @default.
- W3140819097 hasRelatedWork W2597787948 @default.
- W3140819097 hasRelatedWork W2742991909 @default.
- W3140819097 hasRelatedWork W2810865670 @default.
- W3140819097 hasRelatedWork W2907566631 @default.
- W3140819097 hasRelatedWork W2963261224 @default.
- W3140819097 hasRelatedWork W3205394184 @default.
- W3140819097 hasRelatedWork W3210497558 @default.
- W3140819097 hasRelatedWork W4285601942 @default.
- W3140819097 hasRelatedWork W4361193208 @default.
- W3140819097 isParatext "false" @default.
- W3140819097 isRetracted "false" @default.
- W3140819097 magId "3140819097" @default.
- W3140819097 workType "article" @default.