Matches in SemOpenAlex for { <https://semopenalex.org/work/W3206818284> ?p ?o ?g. }
- W3206818284 abstract "Contrastive learning of auditory and visual perception has been extremely successful when investigated individually. However, there are still major questions on how we could integrate principles learned from both domains to attain effective audiovisual representations. In this paper, we present a contrastive framework to learn audiovisual representations from unlabeled videos. The type and strength of augmentations utilized during self-supervised pre-training play a crucial role for contrastive frameworks to work sufficiently. Hence, we extensively investigate composition of temporal augmentations suitable for learning audiovisual representations; we find lossy spatio-temporal transformations that do not corrupt the temporal coherency of videos are the most effective. Furthermore, we show that the effectiveness of these transformations scales with higher temporal resolution and stronger transformation intensity. Compared to self-supervised models pre-trained on only sampling-based temporal augmentation, self-supervised models pre-trained with our temporal augmentations lead to approximately 6.5% gain on linear classifier performance on AVE dataset. Lastly, we show that despite their simplicity, our proposed transformations work well across self-supervised learning frameworks (SimSiam, MoCoV3, etc), and benchmark audiovisual dataset (AVE)." @default.
- W3206818284 created "2021-10-25" @default.
- W3206818284 creator A5057801943 @default.
- W3206818284 creator A5086048817 @default.
- W3206818284 date "2021-10-13" @default.
- W3206818284 modified "2023-09-27" @default.
- W3206818284 title "The Impact of Spatiotemporal Augmentations on Self-Supervised Audiovisual Representation Learning." @default.
- W3206818284 cites W219040644 @default.
- W3206818284 cites W2194775991 @default.
- W3206818284 cites W2308529009 @default.
- W3206818284 cites W2321533354 @default.
- W3206818284 cites W2326925005 @default.
- W3206818284 cites W2511428026 @default.
- W3206818284 cites W2558661413 @default.
- W3206818284 cites W2575671312 @default.
- W3206818284 cites W2619697695 @default.
- W3206818284 cites W2622263826 @default.
- W3206818284 cites W2726515241 @default.
- W3206818284 cites W2842511635 @default.
- W3206818284 cites W2962728572 @default.
- W3206818284 cites W2962742544 @default.
- W3206818284 cites W2962756039 @default.
- W3206818284 cites W2962865004 @default.
- W3206818284 cites W2962960500 @default.
- W3206818284 cites W2963091558 @default.
- W3206818284 cites W2963115079 @default.
- W3206818284 cites W2963218389 @default.
- W3206818284 cites W2963263347 @default.
- W3206818284 cites W2963420272 @default.
- W3206818284 cites W2963426332 @default.
- W3206818284 cites W2963495051 @default.
- W3206818284 cites W2963680395 @default.
- W3206818284 cites W2963801643 @default.
- W3206818284 cites W2964048159 @default.
- W3206818284 cites W2964109005 @default.
- W3206818284 cites W2979579363 @default.
- W3206818284 cites W2981851635 @default.
- W3206818284 cites W2982619606 @default.
- W3206818284 cites W2990467045 @default.
- W3206818284 cites W2990503944 @default.
- W3206818284 cites W2996383576 @default.
- W3206818284 cites W3002552512 @default.
- W3206818284 cites W3010094231 @default.
- W3206818284 cites W3012540512 @default.
- W3206818284 cites W3034978746 @default.
- W3206818284 cites W3035058308 @default.
- W3206818284 cites W3035524453 @default.
- W3206818284 cites W3099782249 @default.
- W3206818284 cites W3101537861 @default.
- W3206818284 cites W3105422445 @default.
- W3206818284 cites W3108655343 @default.
- W3206818284 cites W3134486096 @default.
- W3206818284 cites W3145385912 @default.
- W3206818284 cites W3145450063 @default.
- W3206818284 cites W3159394092 @default.
- W3206818284 cites W3163937874 @default.
- W3206818284 cites W3170837227 @default.
- W3206818284 cites W3171007011 @default.
- W3206818284 cites W343636949 @default.
- W3206818284 hasPublicationYear "2021" @default.
- W3206818284 type Work @default.
- W3206818284 sameAs 3206818284 @default.
- W3206818284 citedByCount "0" @default.
- W3206818284 crossrefType "posted-content" @default.
- W3206818284 hasAuthorship W3206818284A5057801943 @default.
- W3206818284 hasAuthorship W3206818284A5086048817 @default.
- W3206818284 hasConcept C104317684 @default.
- W3206818284 hasConcept C119857082 @default.
- W3206818284 hasConcept C13280743 @default.
- W3206818284 hasConcept C136389625 @default.
- W3206818284 hasConcept C154945302 @default.
- W3206818284 hasConcept C15744967 @default.
- W3206818284 hasConcept C165021410 @default.
- W3206818284 hasConcept C169760540 @default.
- W3206818284 hasConcept C17744445 @default.
- W3206818284 hasConcept C185592680 @default.
- W3206818284 hasConcept C185798385 @default.
- W3206818284 hasConcept C199539241 @default.
- W3206818284 hasConcept C204241405 @default.
- W3206818284 hasConcept C205649164 @default.
- W3206818284 hasConcept C26760741 @default.
- W3206818284 hasConcept C2776359362 @default.
- W3206818284 hasConcept C41008148 @default.
- W3206818284 hasConcept C50644808 @default.
- W3206818284 hasConcept C55493867 @default.
- W3206818284 hasConcept C59404180 @default.
- W3206818284 hasConcept C94625758 @default.
- W3206818284 hasConcept C95623464 @default.
- W3206818284 hasConceptScore W3206818284C104317684 @default.
- W3206818284 hasConceptScore W3206818284C119857082 @default.
- W3206818284 hasConceptScore W3206818284C13280743 @default.
- W3206818284 hasConceptScore W3206818284C136389625 @default.
- W3206818284 hasConceptScore W3206818284C154945302 @default.
- W3206818284 hasConceptScore W3206818284C15744967 @default.
- W3206818284 hasConceptScore W3206818284C165021410 @default.
- W3206818284 hasConceptScore W3206818284C169760540 @default.
- W3206818284 hasConceptScore W3206818284C17744445 @default.
- W3206818284 hasConceptScore W3206818284C185592680 @default.
- W3206818284 hasConceptScore W3206818284C185798385 @default.
- W3206818284 hasConceptScore W3206818284C199539241 @default.