FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3006314575> ?p ?o ?g. }

• W3006314575 abstract "Recent interest in self-supervised dense tracking has yielded rapid progress, but performance still remains far from supervised methods. We propose a dense tracking model trained on videos without any annotations that surpasses previous self-supervised methods on existing benchmarks by a significant margin (+15%), and achieves performance comparable to supervised methods. In this paper, we first reassess the traditional choices used for self-supervised training and reconstruction loss by conducting thorough experiments that finally elucidate the optimal choices. Second, we further improve on existing methods by augmenting our architecture with a crucial memory component. Third, we benchmark on large-scale semi-supervised video object segmentation(aka. dense tracking), and propose a new metric: generalizability. Our first two contributions yield a self-supervised network that for the first time is competitive with supervised methods on standard evaluation metrics of dense tracking. When measuring generalizability, we show self-supervised approaches are actually superior to the majority of supervised methods. We believe this new generalizability metric can better capture the real-world use-cases for dense tracking, and will spur new interest in this research direction." @default.
• W3006314575 created "2020-02-24" @default.
• W3006314575 creator A5076097168 @default.
• W3006314575 creator A5076191170 @default.
• W3006314575 creator A5076947156 @default.
• W3006314575 date "2020-02-18" @default.
• W3006314575 modified "2023-09-26" @default.
• W3006314575 title "MAST: A Memory-Augmented Self-supervised Tracker" @default.
• W3006314575 cites W1514535095 @default.
• W3006314575 cites W1520997877 @default.
• W3006314575 cites W1523714666 @default.
• W3006314575 cites W1586939924 @default.
• W3006314575 cites W1861492603 @default.
• W3006314575 cites W1924770834 @default.
• W3006314575 cites W2032958483 @default.
• W3006314575 cites W2064675550 @default.
• W3006314575 cites W2108598243 @default.
• W3006314575 cites W2129112648 @default.
• W3006314575 cites W2158827467 @default.
• W3006314575 cites W2175760607 @default.
• W3006314575 cites W219040644 @default.
• W3006314575 cites W2194775991 @default.
• W3006314575 cites W2198618282 @default.
• W3006314575 cites W2267126114 @default.
• W3006314575 cites W2326925005 @default.
• W3006314575 cites W2414711238 @default.
• W3006314575 cites W2470139095 @default.
• W3006314575 cites W2487442924 @default.
• W3006314575 cites W2550462002 @default.
• W3006314575 cites W2626778328 @default.
• W3006314575 cites W2758832202 @default.
• W3006314575 cites W2794847483 @default.
• W3006314575 cites W2798271879 @default.
• W3006314575 cites W2799087757 @default.
• W3006314575 cites W2799239273 @default.
• W3006314575 cites W2809334950 @default.
• W3006314575 cites W2884460600 @default.
• W3006314575 cites W2889658408 @default.
• W3006314575 cites W2890447039 @default.
• W3006314575 cites W2896457183 @default.
• W3006314575 cites W2902016181 @default.
• W3006314575 cites W2916743882 @default.
• W3006314575 cites W2916797271 @default.
• W3006314575 cites W2921536280 @default.
• W3006314575 cites W2926645869 @default.
• W3006314575 cites W2927438889 @default.
• W3006314575 cites W2929960531 @default.
• W3006314575 cites W2943695751 @default.
• W3006314575 cites W2949213324 @default.
• W3006314575 cites W2949302831 @default.
• W3006314575 cites W2950527759 @default.
• W3006314575 cites W2952122856 @default.
• W3006314575 cites W2952913664 @default.
• W3006314575 cites W2955874753 @default.
• W3006314575 cites W2962728572 @default.
• W3006314575 cites W2962825871 @default.
• W3006314575 cites W2962942822 @default.
• W3006314575 cites W2963045453 @default.
• W3006314575 cites W2963091558 @default.
• W3006314575 cites W2963227409 @default.
• W3006314575 cites W2963426332 @default.
• W3006314575 cites W2963503215 @default.
• W3006314575 cites W2963732700 @default.
• W3006314575 cites W2963773230 @default.
• W3006314575 cites W2963814513 @default.
• W3006314575 cites W2964037671 @default.
• W3006314575 cites W2964040697 @default.
• W3006314575 cites W2964091467 @default.
• W3006314575 cites W2964135029 @default.
• W3006314575 cites W2964218467 @default.
• W3006314575 cites W2964286567 @default.
• W3006314575 cites W2964308564 @default.
• W3006314575 cites W2964343881 @default.
• W3006314575 cites W2964700958 @default.
• W3006314575 cites W2967622921 @default.
• W3006314575 cites W2970235454 @default.
• W3006314575 cites W2990205821 @default.
• W3006314575 cites W3010874390 @default.
• W3006314575 cites W603908379 @default.
• W3006314575 doi "https://doi.org/10.48550/arxiv.2002.07793" @default.
• W3006314575 hasPublicationYear "2020" @default.
• W3006314575 type Work @default.
• W3006314575 sameAs 3006314575 @default.
• W3006314575 citedByCount "5" @default.
• W3006314575 countsByYear W30063145752020 @default.
• W3006314575 countsByYear W30063145752021 @default.
• W3006314575 crossrefType "posted-content" @default.
• W3006314575 hasAuthorship W3006314575A5076097168 @default.
• W3006314575 hasAuthorship W3006314575A5076191170 @default.
• W3006314575 hasAuthorship W3006314575A5076947156 @default.
• W3006314575 hasBestOaLocation W30063145751 @default.
• W3006314575 hasConcept C105795698 @default.
• W3006314575 hasConcept C119857082 @default.
• W3006314575 hasConcept C13280743 @default.
• W3006314575 hasConcept C154945302 @default.
• W3006314575 hasConcept C162324750 @default.
• W3006314575 hasConcept C176217482 @default.
• W3006314575 hasConcept C185798385 @default.
• W3006314575 hasConcept C205649164 @default.
• W3006314575 hasConcept C21547014 @default.

• next