Matches in SemOpenAlex for { <https://semopenalex.org/work/W4383301864> ?p ?o ?g. }
- W4383301864 endingPage "4" @default.
- W4383301864 startingPage "4" @default.
- W4383301864 abstract "In laboratory object recognition tasks based on undistorted photographs, both adult humans and deep neural networks (DNNs) perform close to ceiling. Unlike adults’, whose object recognition performance is robust against a wide range of image distortions, DNNs trained on standard ImageNet (1.3M images) perform poorly on distorted images. However, the last 2 years have seen impressive gains in DNN distortion robustness, predominantly achieved through ever-increasing large-scale datasets—orders of magnitude larger than ImageNet. Although this simple brute-force approach is very effective in achieving human-level robustness in DNNs, it raises the question of whether human robustness, too, is simply due to extensive experience with (distorted) visual input during childhood and beyond. Here we investigate this question by comparing the core object recognition performance of 146 children (aged 4–15 years) against adults and against DNNs. We find, first, that already 4- to 6-year-olds show remarkable robustness to image distortions and outperform DNNs trained on ImageNet. Second, we estimated the number of images children had been exposed to during their lifetime. Compared with various DNNs, children’s high robustness requires relatively little data. Third, when recognizing objects, children—like adults but unlike DNNs—rely heavily on shape but not on texture cues. Together our results suggest that the remarkable robustness to distortions emerges early in the developmental trajectory of human object recognition and is unlikely the result of a mere accumulation of experience with distorted visual input. Even though current DNNs match human performance regarding robustness, they seem to rely on different and more data-hungry strategies to do so." @default.
- W4383301864 created "2023-07-07" @default.
- W4383301864 creator A5007987869 @default.
- W4383301864 creator A5043113119 @default.
- W4383301864 creator A5043258699 @default.
- W4383301864 date "2023-07-06" @default.
- W4383301864 modified "2023-10-16" @default.
- W4383301864 title "The developmental trajectory of object recognition robustness: Children are like small adults but unlike big deep neural networks" @default.
- W4383301864 cites W1499837978 @default.
- W4383301864 cites W1677182931 @default.
- W4383301864 cites W1887706683 @default.
- W4383301864 cites W1971017968 @default.
- W4383301864 cites W1979623747 @default.
- W4383301864 cites W1981299032 @default.
- W4383301864 cites W1982374456 @default.
- W4383301864 cites W2004995289 @default.
- W4383301864 cites W2014621385 @default.
- W4383301864 cites W2022620441 @default.
- W4383301864 cites W2026942141 @default.
- W4383301864 cites W2030265437 @default.
- W4383301864 cites W2046394129 @default.
- W4383301864 cites W2059179319 @default.
- W4383301864 cites W2059799772 @default.
- W4383301864 cites W2063303346 @default.
- W4383301864 cites W2068824491 @default.
- W4383301864 cites W2070544989 @default.
- W4383301864 cites W2080624334 @default.
- W4383301864 cites W2081580037 @default.
- W4383301864 cites W2092785577 @default.
- W4383301864 cites W2096698697 @default.
- W4383301864 cites W2098216184 @default.
- W4383301864 cites W2100922564 @default.
- W4383301864 cites W2102187890 @default.
- W4383301864 cites W2105824230 @default.
- W4383301864 cites W2107531711 @default.
- W4383301864 cites W2117539524 @default.
- W4383301864 cites W2122000263 @default.
- W4383301864 cites W2138692731 @default.
- W4383301864 cites W2141559645 @default.
- W4383301864 cites W2142209090 @default.
- W4383301864 cites W2144730066 @default.
- W4383301864 cites W2145339207 @default.
- W4383301864 cites W2149386455 @default.
- W4383301864 cites W2152749681 @default.
- W4383301864 cites W2156406284 @default.
- W4383301864 cites W2162950292 @default.
- W4383301864 cites W2167956961 @default.
- W4383301864 cites W2273969094 @default.
- W4383301864 cites W2320331759 @default.
- W4383301864 cites W2553608650 @default.
- W4383301864 cites W2592295070 @default.
- W4383301864 cites W2768701474 @default.
- W4383301864 cites W2790065565 @default.
- W4383301864 cites W2796261340 @default.
- W4383301864 cites W2903867357 @default.
- W4383301864 cites W2913606428 @default.
- W4383301864 cites W2954996726 @default.
- W4383301864 cites W2963305465 @default.
- W4383301864 cites W2971923036 @default.
- W4383301864 cites W3084747096 @default.
- W4383301864 cites W3095147100 @default.
- W4383301864 cites W3119948327 @default.
- W4383301864 cites W3138341448 @default.
- W4383301864 cites W3161502969 @default.
- W4383301864 cites W3176923149 @default.
- W4383301864 cites W3200213953 @default.
- W4383301864 cites W3201124393 @default.
- W4383301864 cites W4228996534 @default.
- W4383301864 cites W4236804248 @default.
- W4383301864 cites W4238109516 @default.
- W4383301864 cites W4239280998 @default.
- W4383301864 cites W4240548534 @default.
- W4383301864 cites W4241027858 @default.
- W4383301864 cites W4241074508 @default.
- W4383301864 cites W4246177170 @default.
- W4383301864 cites W4246194497 @default.
- W4383301864 cites W4254269720 @default.
- W4383301864 cites W4281480780 @default.
- W4383301864 cites W900492034 @default.
- W4383301864 doi "https://doi.org/10.1167/jov.23.7.4" @default.
- W4383301864 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37410494" @default.
- W4383301864 hasPublicationYear "2023" @default.
- W4383301864 type Work @default.
- W4383301864 citedByCount "0" @default.
- W4383301864 crossrefType "journal-article" @default.
- W4383301864 hasAuthorship W4383301864A5007987869 @default.
- W4383301864 hasAuthorship W4383301864A5043113119 @default.
- W4383301864 hasAuthorship W4383301864A5043258699 @default.
- W4383301864 hasBestOaLocation W43833018641 @default.
- W4383301864 hasConcept C104317684 @default.
- W4383301864 hasConcept C153180895 @default.
- W4383301864 hasConcept C154945302 @default.
- W4383301864 hasConcept C2781238097 @default.
- W4383301864 hasConcept C2984842247 @default.
- W4383301864 hasConcept C31972630 @default.
- W4383301864 hasConcept C41008148 @default.
- W4383301864 hasConcept C50644808 @default.
- W4383301864 hasConcept C55493867 @default.