FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3048787757> ?p ?o ?g. }

• W3048787757 abstract "Simulation is increasingly being used for generating large labelled datasets in many machine learning problems. Recent methods have focused on adjusting simulator parameters with the goal of maximising accuracy on a validation task, usually relying on REINFORCE-like gradient estimators. However these approaches are very expensive as they treat the entire data generation, model training, and validation pipeline as a black-box and require multiple costly objective evaluations at each iteration. We propose an efficient alternative for optimal synthetic data generation, based on a novel differentiable approximation of the objective. This allows us to optimize the simulator, which may be non-differentiable, requiring only one objective evaluation at each iteration with a little overhead. We demonstrate on a state-of-the-art photorealistic renderer that the proposed method finds the optimal data distribution faster (up to $50times$), with significantly reduced training data generation (up to $30times$) and better accuracy ($+8.7%$) on real-world test datasets than previous methods." @default.
• W3048787757 created "2020-08-18" @default.
• W3048787757 creator A5042899882 @default.
• W3048787757 creator A5045147286 @default.
• W3048787757 creator A5057080001 @default.
• W3048787757 creator A5063994783 @default.
• W3048787757 creator A5081255348 @default.
• W3048787757 date "2020-08-16" @default.
• W3048787757 modified "2023-09-27" @default.
• W3048787757 title "AutoSimulate: (Quickly) Learning Synthetic Data Generation" @default.
• W3048787757 cites W1591870335 @default.
• W3048787757 cites W1646152356 @default.
• W3048787757 cites W1901129140 @default.
• W3048787757 cites W1902237438 @default.
• W3048787757 cites W1903029394 @default.
• W3048787757 cites W2097571405 @default.
• W3048787757 cites W2098841537 @default.
• W3048787757 cites W2101234009 @default.
• W3048787757 cites W2119717200 @default.
• W3048787757 cites W2124659975 @default.
• W3048787757 cites W2131241448 @default.
• W3048787757 cites W2156417353 @default.
• W3048787757 cites W2236623899 @default.
• W3048787757 cites W2341569833 @default.
• W3048787757 cites W2395611524 @default.
• W3048787757 cites W2397830550 @default.
• W3048787757 cites W2431874326 @default.
• W3048787757 cites W2487365028 @default.
• W3048787757 cites W2561715562 @default.
• W3048787757 cites W2563100679 @default.
• W3048787757 cites W2570343428 @default.
• W3048787757 cites W2576289912 @default.
• W3048787757 cites W2604236302 @default.
• W3048787757 cites W2604763608 @default.
• W3048787757 cites W2768879211 @default.
• W3048787757 cites W2883820570 @default.
• W3048787757 cites W2885656008 @default.
• W3048787757 cites W2888752296 @default.
• W3048787757 cites W2907670343 @default.
• W3048787757 cites W2949907962 @default.
• W3048787757 cites W2954637159 @default.
• W3048787757 cites W2962781217 @default.
• W3048787757 cites W2962836306 @default.
• W3048787757 cites W2962897886 @default.
• W3048787757 cites W2963150697 @default.
• W3048787757 cites W2963188159 @default.
• W3048787757 cites W2963233958 @default.
• W3048787757 cites W2963271314 @default.
• W3048787757 cites W2963306862 @default.
• W3048787757 cites W2963630234 @default.
• W3048787757 cites W2963642948 @default.
• W3048787757 cites W2963675327 @default.
• W3048787757 cites W2963742597 @default.
• W3048787757 cites W2964047820 @default.
• W3048787757 cites W2964125128 @default.
• W3048787757 cites W2964271185 @default.
• W3048787757 cites W2970399728 @default.
• W3048787757 cites W2971217834 @default.
• W3048787757 cites W2985936292 @default.
• W3048787757 cites W2990675203 @default.
• W3048787757 cites W639708223 @default.
• W3048787757 cites W764651262 @default.
• W3048787757 hasPublicationYear "2020" @default.
• W3048787757 type Work @default.
• W3048787757 sameAs 3048787757 @default.
• W3048787757 citedByCount "0" @default.
• W3048787757 crossrefType "posted-content" @default.
• W3048787757 hasAuthorship W3048787757A5042899882 @default.
• W3048787757 hasAuthorship W3048787757A5045147286 @default.
• W3048787757 hasAuthorship W3048787757A5057080001 @default.
• W3048787757 hasAuthorship W3048787757A5063994783 @default.
• W3048787757 hasAuthorship W3048787757A5081255348 @default.
• W3048787757 hasConcept C105795698 @default.
• W3048787757 hasConcept C111919701 @default.
• W3048787757 hasConcept C119857082 @default.
• W3048787757 hasConcept C124101348 @default.
• W3048787757 hasConcept C134306372 @default.
• W3048787757 hasConcept C154945302 @default.
• W3048787757 hasConcept C160920958 @default.
• W3048787757 hasConcept C162324750 @default.
• W3048787757 hasConcept C185429906 @default.
• W3048787757 hasConcept C187736073 @default.
• W3048787757 hasConcept C199360897 @default.
• W3048787757 hasConcept C202615002 @default.
• W3048787757 hasConcept C2779960059 @default.
• W3048787757 hasConcept C2780451532 @default.
• W3048787757 hasConcept C33923547 @default.
• W3048787757 hasConcept C41008148 @default.
• W3048787757 hasConcept C43521106 @default.
• W3048787757 hasConceptScore W3048787757C105795698 @default.
• W3048787757 hasConceptScore W3048787757C111919701 @default.
• W3048787757 hasConceptScore W3048787757C119857082 @default.
• W3048787757 hasConceptScore W3048787757C124101348 @default.
• W3048787757 hasConceptScore W3048787757C134306372 @default.
• W3048787757 hasConceptScore W3048787757C154945302 @default.
• W3048787757 hasConceptScore W3048787757C160920958 @default.
• W3048787757 hasConceptScore W3048787757C162324750 @default.
• W3048787757 hasConceptScore W3048787757C185429906 @default.
• W3048787757 hasConceptScore W3048787757C187736073 @default.
• W3048787757 hasConceptScore W3048787757C199360897 @default.

• next