Matches in SemOpenAlex for { <https://semopenalex.org/work/W4286584259> ?p ?o ?g. }
Showing items 1 to 78 of
78
with 100 items per page.
- W4286584259 abstract "What is the best way to learn a universal face representation? Recent work on Deep Learning in the area of face analysis has focused on supervised learning for specific tasks of interest (e.g. face recognition, facial landmark localization etc.) but has overlooked the overarching question of how to find a facial representation that can be readily adapted to several facial analysis tasks and datasets. To this end, we make the following 4 contributions: (a) we introduce, for the first time, a comprehensive evaluation benchmark for facial representation learning consisting of 5 important face analysis tasks. (b) We systematically investigate two ways of large-scale representation learning applied to faces: supervised and unsupervised pre-training. Importantly, we focus our evaluations on the case of few-shot facial learning. (c) We investigate important properties of the training datasets including their size and quality (labelled, unlabelled or even uncurated). (d) To draw our conclusions, we conducted a very large number of experiments. Our main two findings are: (1) Unsupervised pre-training on completely in-the-wild, uncurated data provides consistent and, in some cases, significant accuracy improvements for all facial tasks considered. (2) Many existing facial video datasets seem to have a large amount of redundancy. We will release code, and pre-trained models to facilitate future research." @default.
- W4286584259 created "2022-07-22" @default.
- W4286584259 creator A5011737734 @default.
- W4286584259 creator A5017072420 @default.
- W4286584259 creator A5024224610 @default.
- W4286584259 creator A5028453074 @default.
- W4286584259 creator A5072822225 @default.
- W4286584259 creator A5089860286 @default.
- W4286584259 date "2021-03-30" @default.
- W4286584259 modified "2023-09-24" @default.
- W4286584259 title "Pre-training strategies and datasets for facial representation learning" @default.
- W4286584259 doi "https://doi.org/10.48550/arxiv.2103.16554" @default.
- W4286584259 hasPublicationYear "2021" @default.
- W4286584259 type Work @default.
- W4286584259 citedByCount "0" @default.
- W4286584259 crossrefType "posted-content" @default.
- W4286584259 hasAuthorship W4286584259A5011737734 @default.
- W4286584259 hasAuthorship W4286584259A5017072420 @default.
- W4286584259 hasAuthorship W4286584259A5024224610 @default.
- W4286584259 hasAuthorship W4286584259A5028453074 @default.
- W4286584259 hasAuthorship W4286584259A5072822225 @default.
- W4286584259 hasAuthorship W4286584259A5089860286 @default.
- W4286584259 hasBestOaLocation W42865842591 @default.
- W4286584259 hasConcept C111919701 @default.
- W4286584259 hasConcept C119857082 @default.
- W4286584259 hasConcept C13280743 @default.
- W4286584259 hasConcept C144024400 @default.
- W4286584259 hasConcept C152124472 @default.
- W4286584259 hasConcept C153180895 @default.
- W4286584259 hasConcept C154945302 @default.
- W4286584259 hasConcept C17744445 @default.
- W4286584259 hasConcept C185798385 @default.
- W4286584259 hasConcept C199539241 @default.
- W4286584259 hasConcept C205649164 @default.
- W4286584259 hasConcept C2776359362 @default.
- W4286584259 hasConcept C2779304628 @default.
- W4286584259 hasConcept C2780297707 @default.
- W4286584259 hasConcept C36289849 @default.
- W4286584259 hasConcept C41008148 @default.
- W4286584259 hasConcept C59404180 @default.
- W4286584259 hasConcept C8038995 @default.
- W4286584259 hasConcept C94625758 @default.
- W4286584259 hasConceptScore W4286584259C111919701 @default.
- W4286584259 hasConceptScore W4286584259C119857082 @default.
- W4286584259 hasConceptScore W4286584259C13280743 @default.
- W4286584259 hasConceptScore W4286584259C144024400 @default.
- W4286584259 hasConceptScore W4286584259C152124472 @default.
- W4286584259 hasConceptScore W4286584259C153180895 @default.
- W4286584259 hasConceptScore W4286584259C154945302 @default.
- W4286584259 hasConceptScore W4286584259C17744445 @default.
- W4286584259 hasConceptScore W4286584259C185798385 @default.
- W4286584259 hasConceptScore W4286584259C199539241 @default.
- W4286584259 hasConceptScore W4286584259C205649164 @default.
- W4286584259 hasConceptScore W4286584259C2776359362 @default.
- W4286584259 hasConceptScore W4286584259C2779304628 @default.
- W4286584259 hasConceptScore W4286584259C2780297707 @default.
- W4286584259 hasConceptScore W4286584259C36289849 @default.
- W4286584259 hasConceptScore W4286584259C41008148 @default.
- W4286584259 hasConceptScore W4286584259C59404180 @default.
- W4286584259 hasConceptScore W4286584259C8038995 @default.
- W4286584259 hasConceptScore W4286584259C94625758 @default.
- W4286584259 hasLocation W42865842591 @default.
- W4286584259 hasLocation W42865842592 @default.
- W4286584259 hasOpenAccess W4286584259 @default.
- W4286584259 hasPrimaryLocation W42865842591 @default.
- W4286584259 hasRelatedWork W2592385986 @default.
- W4286584259 hasRelatedWork W2608244370 @default.
- W4286584259 hasRelatedWork W2695951553 @default.
- W4286584259 hasRelatedWork W2775464024 @default.
- W4286584259 hasRelatedWork W2899683012 @default.
- W4286584259 hasRelatedWork W2902482704 @default.
- W4286584259 hasRelatedWork W2908875379 @default.
- W4286584259 hasRelatedWork W2970216048 @default.
- W4286584259 hasRelatedWork W4221136938 @default.
- W4286584259 hasRelatedWork W4310034804 @default.
- W4286584259 isParatext "false" @default.
- W4286584259 isRetracted "false" @default.
- W4286584259 workType "article" @default.