Transmix: Attend to mix for vision transformers

JN Chen, S Sun, J He, PHS Torr… - Proceedings of the …, 2022 - openaccess.thecvf.com
Mixup-based augmentation has been found to be effective for generalizing models during
training, especially for Vision Transformers (ViTs) since they can easily overfit. However …
被引用次数：92 相关文章
[PDF] jhu.edu

[PDF][PDF] TransMix: Attend to Mix for Vision Transformers

JN Chen, S Sun, J He, P Torr, A Yuille, S Bai - cs.jhu.edu
Mixup-based augmentation has been found to be effective for generalizing models during
training, especially for Vision Transformers (ViTs) since they can easily overfit. However …

TransMix: Attend to Mix for Vision Transformers

JN Chen, S Sun, J He, P Torr, A Yuille… - 2022 IEEE/CVF …, 2022 - computer.org
Mixup-based augmentation has been found to be effective for generalizing models during
training, especially for Vision Transformers (ViTs) since they can easily overfit. However …

TransMix: Attend to Mix for Vision Transformers

JN Chen, S Sun, J He, P Torr, A Yuille, S Bai - arXiv preprint arXiv …, 2021 - arxiv.org
Mixup-based augmentation has been found to be effective for generalizing models during
training, especially for Vision Transformers (ViTs) since they can easily overfit. However …

TransMix: attend to mix for Vision Transformers

JN Chen, S Sun, J He, P Torr, A Yuille, S Bai - 2022 - ora.ox.ac.uk
Mixup-based augmentation has been found to be effective for generalizing models during
training, especially for Vision Transformers (ViTs) since they can easily overfit. However …

TransMix: Attend to Mix for Vision Transformers

JN Chen, S Sun, J He, P Torr, A Yuille, S Bai - arXiv e-prints, 2021 - ui.adsabs.harvard.edu
Mixup-based augmentation has been found to be effective for generalizing models during
training, especially for Vision Transformers (ViTs) since they can easily overfit. However …

TransMix: Attend to Mix for Vision Transformers

JN Chen, S Sun, J He, P Torr… - 2022 IEEE/CVF …, 2022 - ieeexplore.ieee.org
Mixup-based augmentation has been found to be effective for generalizing models during
training, especially for Vision Transformers (ViTs) since they can easily overfit. However …