Matches in SemOpenAlex for { <https://semopenalex.org/work/W3157916917> ?p ?o ?g. }
- W3157916917 abstract "We present a multimodal framework to learn general audio representations from videos. Existing contrastive audio representation learning methods mainly focus on using the audio modality alone during training. In this work, we show that additional information contained in video can be utilized to greatly improve the learned features. First, we demonstrate that our contrastive framework does not require high resolution images to learn good audio features. This allows us to scale up the training batch size, while keeping the computational load incurred by the additional video modality to a reasonable level. Second, we use augmentations that mix together different samples. We show that this is effective to make the proxy task harder, which leads to substantial performance improvements when increasing the batch size. As a result, our audio model achieves a state-of-the-art of 42.4 mAP on the AudioSet classification downstream task, closing the gap between supervised and self-supervised methods trained on the same dataset. Moreover, we show that our method is advantageous on a broad range of non-semantic audio tasks, including speaker identification, keyword spotting, language identification, and music instrument classification." @default.
- W3157916917 created "2021-05-10" @default.
- W3157916917 creator A5012826868 @default.
- W3157916917 creator A5018641754 @default.
- W3157916917 creator A5020406406 @default.
- W3157916917 creator A5030447052 @default.
- W3157916917 creator A5039776447 @default.
- W3157916917 date "2021-04-26" @default.
- W3157916917 modified "2023-09-27" @default.
- W3157916917 title "Multimodal Self-Supervised Learning of General Audio Representations." @default.
- W3157916917 cites W1494198834 @default.
- W3157916917 cites W2127589108 @default.
- W3157916917 cites W2171590421 @default.
- W3157916917 cites W2219249508 @default.
- W3157916917 cites W2526050071 @default.
- W3157916917 cites W2593116425 @default.
- W3157916917 cites W2606176153 @default.
- W3157916917 cites W2619697695 @default.
- W3157916917 cites W2726515241 @default.
- W3157916917 cites W2767754137 @default.
- W3157916917 cites W2797583228 @default.
- W3157916917 cites W2842511635 @default.
- W3157916917 cites W2883595988 @default.
- W3157916917 cites W2944828972 @default.
- W3157916917 cites W2955425717 @default.
- W3157916917 cites W2963263347 @default.
- W3157916917 cites W2963341956 @default.
- W3157916917 cites W2963399829 @default.
- W3157916917 cites W2964121744 @default.
- W3157916917 cites W2971155163 @default.
- W3157916917 cites W2973049979 @default.
- W3157916917 cites W2990152177 @default.
- W3157916917 cites W3006926732 @default.
- W3157916917 cites W3015817524 @default.
- W3157916917 cites W3015949486 @default.
- W3157916917 cites W3034978746 @default.
- W3157916917 cites W3035058308 @default.
- W3157916917 cites W3035524453 @default.
- W3157916917 cites W3035635319 @default.
- W3157916917 cites W3093563057 @default.
- W3157916917 cites W3094550259 @default.
- W3157916917 cites W3095727342 @default.
- W3157916917 cites W3099782249 @default.
- W3157916917 cites W3100177202 @default.
- W3157916917 cites W3102342027 @default.
- W3157916917 cites W3104805957 @default.
- W3157916917 cites W3124414150 @default.
- W3157916917 cites W3134486096 @default.
- W3157916917 cites W3135367836 @default.
- W3157916917 cites W3210683364 @default.
- W3157916917 hasPublicationYear "2021" @default.
- W3157916917 type Work @default.
- W3157916917 sameAs 3157916917 @default.
- W3157916917 citedByCount "1" @default.
- W3157916917 countsByYear W31579169172021 @default.
- W3157916917 crossrefType "posted-content" @default.
- W3157916917 hasAuthorship W3157916917A5012826868 @default.
- W3157916917 hasAuthorship W3157916917A5018641754 @default.
- W3157916917 hasAuthorship W3157916917A5020406406 @default.
- W3157916917 hasAuthorship W3157916917A5030447052 @default.
- W3157916917 hasAuthorship W3157916917A5039776447 @default.
- W3157916917 hasConcept C119857082 @default.
- W3157916917 hasConcept C120665830 @default.
- W3157916917 hasConcept C121332964 @default.
- W3157916917 hasConcept C154945302 @default.
- W3157916917 hasConcept C162324750 @default.
- W3157916917 hasConcept C17744445 @default.
- W3157916917 hasConcept C187736073 @default.
- W3157916917 hasConcept C192209626 @default.
- W3157916917 hasConcept C199539241 @default.
- W3157916917 hasConcept C204321447 @default.
- W3157916917 hasConcept C2778775528 @default.
- W3157916917 hasConcept C2780226545 @default.
- W3157916917 hasConcept C2780451532 @default.
- W3157916917 hasConcept C2781213101 @default.
- W3157916917 hasConcept C28490314 @default.
- W3157916917 hasConcept C41008148 @default.
- W3157916917 hasConceptScore W3157916917C119857082 @default.
- W3157916917 hasConceptScore W3157916917C120665830 @default.
- W3157916917 hasConceptScore W3157916917C121332964 @default.
- W3157916917 hasConceptScore W3157916917C154945302 @default.
- W3157916917 hasConceptScore W3157916917C162324750 @default.
- W3157916917 hasConceptScore W3157916917C17744445 @default.
- W3157916917 hasConceptScore W3157916917C187736073 @default.
- W3157916917 hasConceptScore W3157916917C192209626 @default.
- W3157916917 hasConceptScore W3157916917C199539241 @default.
- W3157916917 hasConceptScore W3157916917C204321447 @default.
- W3157916917 hasConceptScore W3157916917C2778775528 @default.
- W3157916917 hasConceptScore W3157916917C2780226545 @default.
- W3157916917 hasConceptScore W3157916917C2780451532 @default.
- W3157916917 hasConceptScore W3157916917C2781213101 @default.
- W3157916917 hasConceptScore W3157916917C28490314 @default.
- W3157916917 hasConceptScore W3157916917C41008148 @default.
- W3157916917 hasLocation W31579169171 @default.
- W3157916917 hasOpenAccess W3157916917 @default.
- W3157916917 hasPrimaryLocation W31579169171 @default.
- W3157916917 hasRelatedWork W2184188583 @default.
- W3157916917 hasRelatedWork W2888678380 @default.
- W3157916917 hasRelatedWork W2890052321 @default.
- W3157916917 hasRelatedWork W2931316642 @default.