Zobrazeno 1 - 10
of 246
pro vyhledávání: '"Amir Zamir"'
Publikováno v:
Lecture Notes in Computer Science ISBN: 9783031198359
We propose a pre-training strategy called Multi-modal Multi-task Masked Autoencoders (MultiMAE). It differs from standard Masked Autoencoding in two key aspects: I) it can optionally accept additional modalities of information in the input besides th
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::d8a7d0dc7e00cbb1444d08e078f01491
https://doi.org/10.1007/978-3-031-19836-6_20
https://doi.org/10.1007/978-3-031-19836-6_20
Publikováno v:
Pakistan Engineering Review; 9/16/2023, Vol. 48 Issue 18, p1-15, 15p
Autor:
Luo, Jiayun1 (AUTHOR) jiayun.luo@ntu.edu.sg, Li, Boyang1 (AUTHOR) boyang.li@ntu.edu.sg, Leung, Cyril2 (AUTHOR) CLeung@ntu.edu.sg
Publikováno v:
ACM Computing Surveys. May2024, Vol. 56 Issue 5, p1-39. 39p.
Autor:
MIN, BONAN1 bonanmin@amazon.com, ROSS, HAYLEY2 hayleyross@g.harvard.edu, SULEM, ELIOR3 eliors@seas.upenn.edu, BEN VEYSEH, AMIR POURAN4 apouran@cs.uoregon.edu, THIEN HUU NGUYEN4 thien@cs.uoregon.edu, SAINZ, OSCAR5 oscar.sainz@ehu.eus, AGIRRE, ENEKO5 e.agirre@ehu.eus, HEINTZ, ILANA6 ilana@synopticengineering.com, ROTH, DAN3 danroth@seas.upenn.edu
Publikováno v:
ACM Computing Surveys. Feb2024, Vol. 56 Issue 2, p1-40. 40p.
Publikováno v:
Amir Zamir
We introduce a set of image transformations that can be used as corruptions to evaluate the robustness of models as well as data augmentation mechanisms for training neural networks. The primary distinction of the proposed transformations is that, un
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::d5ee5f3800813da605e784000b904001
https://infoscience.epfl.ch/record/299218
https://infoscience.epfl.ch/record/299218
Autor:
Wang, Xu, Yu, Hongwei, Meng, Xiangxin, Cao, Hongliang, Zhang, Hongyu, Sun, Hailong, Liu, Xudong, Hu, Chunming
Publikováno v:
ACM Transactions on Software Engineering & Methodology; Jul2024, Vol. 33 Issue 6, p1-31, 31p
Publikováno v:
Pakistan Engineering Review; 9/15/2023, Vol. 48 Issue 17, p1-17, 17p
We present a method for making neural network predictions robust to shifts from the training data distribution. The proposed method is based on making predictions via a diverse set of cues (called 'middle domains') and ensembling them into one strong
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::300205c473abd28d907e1913942848be
https://infoscience.epfl.ch/record/294856
https://infoscience.epfl.ch/record/294856
Autor:
Careaga, Chris, Aksoy, Yağız
Publikováno v:
ACM Transactions on Graphics; Feb2024, Vol. 43 Issue 1, p1-24, 24p
Publikováno v:
ACM Transactions on Graphics; Dec2023, Vol. 42 Issue 6, p1-13, 13p