Matches in SemOpenAlex for { <https://semopenalex.org/work/W3038470387> ?p ?o ?g. }
- W3038470387 endingPage "7667" @default.
- W3038470387 startingPage "7656" @default.
- W3038470387 abstract "We propose a novel mixture-of-experts class to optimize computer vision models in accordance with data transfer limitations at test time. Our approach postulates that the minimum acceptable amount of data allowing for highly-accurate results can vary for different input space partitions. Therefore, we consider mixtures where experts require different amounts of data, and train a sparse gating function to divide the input space for each expert. By appropriate hyperparameter selection, our approach is able to bias mixtures of experts towards selecting specific experts over others. In this way, we show that the data transfer optimization between visual sensing and processing can be solved as a convex optimization problem.To demonstrate the relation between data availability and performance, we evaluate biased mixtures on a range of mainstream computer vision problems, namely: (i) single shot detection, (ii) image super resolution, and (iii) realtime video action classification. For all cases, and when experts constitute modified baselines to meet different limits on allowed data utility, biased mixtures significantly outperform previous work optimized to meet the same constraints on available data." @default.
- W3038470387 created "2020-07-10" @default.
- W3038470387 creator A5008047912 @default.
- W3038470387 creator A5083380516 @default.
- W3038470387 date "2020-01-01" @default.
- W3038470387 modified "2023-10-18" @default.
- W3038470387 title "Biased Mixtures of Experts: Enabling Computer Vision Inference Under Data Transfer Limitations" @default.
- W3038470387 cites W1530955034 @default.
- W3038470387 cites W1549358575 @default.
- W3038470387 cites W1580157800 @default.
- W3038470387 cites W1686810756 @default.
- W3038470387 cites W1861492603 @default.
- W3038470387 cites W2008531776 @default.
- W3038470387 cites W2042970394 @default.
- W3038470387 cites W2099471712 @default.
- W3038470387 cites W2100495367 @default.
- W3038470387 cites W2134053106 @default.
- W3038470387 cites W2173520492 @default.
- W3038470387 cites W2179423374 @default.
- W3038470387 cites W2187089797 @default.
- W3038470387 cites W2194775991 @default.
- W3038470387 cites W2242818861 @default.
- W3038470387 cites W2296319761 @default.
- W3038470387 cites W2342662179 @default.
- W3038470387 cites W2346257166 @default.
- W3038470387 cites W24089286 @default.
- W3038470387 cites W2476548250 @default.
- W3038470387 cites W2503339013 @default.
- W3038470387 cites W2519430864 @default.
- W3038470387 cites W2552465432 @default.
- W3038470387 cites W2557728737 @default.
- W3038470387 cites W2579277680 @default.
- W3038470387 cites W2593768305 @default.
- W3038470387 cites W2605488490 @default.
- W3038470387 cites W2608501919 @default.
- W3038470387 cites W2612445135 @default.
- W3038470387 cites W2739757502 @default.
- W3038470387 cites W2741137940 @default.
- W3038470387 cites W2752037867 @default.
- W3038470387 cites W2757028014 @default.
- W3038470387 cites W2778191445 @default.
- W3038470387 cites W2782776028 @default.
- W3038470387 cites W2795155917 @default.
- W3038470387 cites W2797925891 @default.
- W3038470387 cites W2886397856 @default.
- W3038470387 cites W2892278106 @default.
- W3038470387 cites W2894234546 @default.
- W3038470387 cites W2952339051 @default.
- W3038470387 cites W2953318193 @default.
- W3038470387 cites W2962821904 @default.
- W3038470387 cites W2963524571 @default.
- W3038470387 cites W2963636093 @default.
- W3038470387 cites W2963674932 @default.
- W3038470387 cites W2963880930 @default.
- W3038470387 cites W2964122153 @default.
- W3038470387 cites W2964299589 @default.
- W3038470387 cites W2965631471 @default.
- W3038470387 cites W3106250896 @default.
- W3038470387 cites W648143168 @default.
- W3038470387 cites W98520975 @default.
- W3038470387 doi "https://doi.org/10.1109/tip.2020.3005508" @default.
- W3038470387 hasPublicationYear "2020" @default.
- W3038470387 type Work @default.
- W3038470387 sameAs 3038470387 @default.
- W3038470387 citedByCount "0" @default.
- W3038470387 crossrefType "journal-article" @default.
- W3038470387 hasAuthorship W3038470387A5008047912 @default.
- W3038470387 hasAuthorship W3038470387A5083380516 @default.
- W3038470387 hasBestOaLocation W30384703872 @default.
- W3038470387 hasConcept C111919701 @default.
- W3038470387 hasConcept C112680207 @default.
- W3038470387 hasConcept C119857082 @default.
- W3038470387 hasConcept C124101348 @default.
- W3038470387 hasConcept C127705205 @default.
- W3038470387 hasConcept C150899416 @default.
- W3038470387 hasConcept C153180895 @default.
- W3038470387 hasConcept C154945302 @default.
- W3038470387 hasConcept C157972887 @default.
- W3038470387 hasConcept C159985019 @default.
- W3038470387 hasConcept C192562407 @default.
- W3038470387 hasConcept C204323151 @default.
- W3038470387 hasConcept C2524010 @default.
- W3038470387 hasConcept C2776214188 @default.
- W3038470387 hasConcept C33923547 @default.
- W3038470387 hasConcept C41008148 @default.
- W3038470387 hasConcept C8642999 @default.
- W3038470387 hasConceptScore W3038470387C111919701 @default.
- W3038470387 hasConceptScore W3038470387C112680207 @default.
- W3038470387 hasConceptScore W3038470387C119857082 @default.
- W3038470387 hasConceptScore W3038470387C124101348 @default.
- W3038470387 hasConceptScore W3038470387C127705205 @default.
- W3038470387 hasConceptScore W3038470387C150899416 @default.
- W3038470387 hasConceptScore W3038470387C153180895 @default.
- W3038470387 hasConceptScore W3038470387C154945302 @default.
- W3038470387 hasConceptScore W3038470387C157972887 @default.
- W3038470387 hasConceptScore W3038470387C159985019 @default.
- W3038470387 hasConceptScore W3038470387C192562407 @default.
- W3038470387 hasConceptScore W3038470387C204323151 @default.