Shortcuts

EVA

摘要

We launch EVA, a vision-centric foundation model to explore the limits of visual representation at scale using only publicly accessible data. EVA is a vanilla ViT pre-trained to reconstruct the masked out image-text aligned vision features conditioned on visible image patches. Via this pretext task, we can efficiently scale up EVA to one billion parameters, and sets new records on a broad range of representative vision downstream tasks, such as image recognition, video action recognition, object detection, instance segmentation and semantic segmentation without heavy supervised training. Moreover, we observe quantitative changes in scaling EVA result in qualitative changes in transfer learning performance that are not present in other models. For instance, EVA takes a great leap in the challenging large vocabulary instance segmentation task: our model achieves almost the same state-of-the-art performance on LVISv1.0 dataset with over a thousand categories and COCO dataset with only eighty categories. Beyond a pure vision encoder, EVA can also serve as a vision-centric, multi-modal pivot to connect images and text. We find initializing the vision tower of a giant CLIP from EVA can greatly stabilize the training and outperform the training from scratch counterpart with much fewer samples and less compute, providing a new direction for scaling up and accelerating the costly training of multi-modal foundation models.

使用方式

from mmpretrain import inference_model

predict = inference_model('vit-base-p16_eva-mae-style-pre_8xb128-coslr-100e_in1k', 'demo/bird.JPEG')
print(predict['pred_class'])
print(predict['pred_score'])

Models and results

Pretrained models

模型

Params (M)

Flops (G)

配置文件

下载

eva-mae-style_vit-base-p16_16xb256-coslr-400e_in1k

111.78

17.58

config

model | log

beit-l-p14_3rdparty-eva_in21k*

303.18

81.08

config

model

beit-l-p14_eva-pre_3rdparty_in21k*

303.18

81.08

config

model

beit-g-p16_3rdparty-eva_30m*

1011.32

203.52

config

model

beit-g-p14_3rdparty-eva_30m*

1011.60

267.17

config

model

beit-g-p14_eva-30m-pre_3rdparty_in21k*

1011.60

267.17

config

model

Models with * are converted from the official repo. The config files of these models are only for inference. We haven’t reproduce the training results.

Image Classification on ImageNet-1k

模型

预训练

Params (M)

Flops (G)

Top-1 (%)

Top-5 (%)

配置文件

下载

vit-base-p16_eva-mae-style-pre_8xb128-coslr-100e_in1k

EVA MAE STYLE

86.57

17.58

83.70

N/A

config

model | log

vit-base-p16_eva-mae-style-pre_8xb2048-linear-coslr-100e_in1k

EVA MAE STYLE

86.57

17.58

69.00

N/A

config

model | log

beit-l-p14_eva-pre_3rdparty_in1k-196px*

EVA

304.14

61.57

87.94

98.5

config

model

beit-l-p14_eva-in21k-pre_3rdparty_in1k-196px*

EVA ImageNet-21k

304.14

61.57

88.58

98.65

config

model

beit-l-p14_eva-pre_3rdparty_in1k-336px*

EVA

304.53

191.10

88.66

98.75

config

model

beit-l-p14_eva-in21k-pre_3rdparty_in1k-336px*

EVA ImageNet-21k

304.53

191.10

89.17

98.86

config

model

beit-g-p14_eva-30m-in21k-pre_3rdparty_in1k-336px*

EVA 30M ImageNet-21k

1013.01

620.64

89.61

98.93

config

model

beit-g-p14_eva-30m-in21k-pre_3rdparty_in1k-560px*

EVA 30M ImageNet-21k

1014.45

1906.76

89.71

98.96

config

model

Models with * are converted from the official repo. The config files of these models are only for inference. We haven’t reproduce the training results.

引用

@article{EVA,
  title={EVA: Exploring the Limits of Masked Visual Representation Learning at Scale},
  author={Fang, Yuxin and Wang, Wen and Xie, Binhui and Sun, Quan and Wu, Ledell and Wang, Xinggang and Huang, Tiejun and Wang, Xinlong and Cao, Yue},
  journal={arXiv preprint arXiv:2211.07636},
  year={2022}
}
Read the Docs v: latest
Versions
latest
stable
mmcls-1.x
mmcls-0.x
dev
Downloads
epub
On Read the Docs
Project Home
Builds

Free document hosting provided by Read the Docs.