EVA-02: A Visual Representation for Neon Genesis

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Fang, Yuxin, Sun, Quan, Wang, Xinggang, Huang, Tiejun, Wang, Xinlong, Cao, Yue
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909275504246784
author Fang, Yuxin
Sun, Quan
Wang, Xinggang
Huang, Tiejun
Wang, Xinlong
Cao, Yue
author_facet Fang, Yuxin
Sun, Quan
Wang, Xinggang
Huang, Tiejun
Wang, Xinlong
Cao, Yue
contents We launch EVA-02, a next-generation Transformer-based visual representation pre-trained to reconstruct strong and robust language-aligned vision features via masked image modeling. With an updated plain Transformer architecture as well as extensive pre-training from an open & accessible giant CLIP vision encoder, EVA-02 demonstrates superior performance compared to prior state-of-the-art approaches across various representative vision tasks, while utilizing significantly fewer parameters and compute budgets. Notably, using exclusively publicly accessible training data, EVA-02 with only 304M parameters achieves a phenomenal 90.0 fine-tuning top-1 accuracy on ImageNet-1K val set. Additionally, our EVA-02-CLIP can reach up to 80.4 zero-shot top-1 on ImageNet-1K, outperforming the previous largest & best open-sourced CLIP with only ~1/6 parameters and ~1/6 image-text training data. We offer four EVA-02 variants in various model sizes, ranging from 6M to 304M parameters, all with impressive performance. To facilitate open access and open research, we release the complete suite of EVA-02 to the community at https://github.com/baaivision/EVA/tree/master/EVA-02.
format Preprint
id arxiv_https___arxiv_org_abs_2303_11331
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle EVA-02: A Visual Representation for Neon Genesis
Fang, Yuxin
Sun, Quan
Wang, Xinggang
Huang, Tiejun
Wang, Xinlong
Cao, Yue
Computer Vision and Pattern Recognition
Computation and Language
We launch EVA-02, a next-generation Transformer-based visual representation pre-trained to reconstruct strong and robust language-aligned vision features via masked image modeling. With an updated plain Transformer architecture as well as extensive pre-training from an open & accessible giant CLIP vision encoder, EVA-02 demonstrates superior performance compared to prior state-of-the-art approaches across various representative vision tasks, while utilizing significantly fewer parameters and compute budgets. Notably, using exclusively publicly accessible training data, EVA-02 with only 304M parameters achieves a phenomenal 90.0 fine-tuning top-1 accuracy on ImageNet-1K val set. Additionally, our EVA-02-CLIP can reach up to 80.4 zero-shot top-1 on ImageNet-1K, outperforming the previous largest & best open-sourced CLIP with only ~1/6 parameters and ~1/6 image-text training data. We offer four EVA-02 variants in various model sizes, ranging from 6M to 304M parameters, all with impressive performance. To facilitate open access and open research, we release the complete suite of EVA-02 to the community at https://github.com/baaivision/EVA/tree/master/EVA-02.
title EVA-02: A Visual Representation for Neon Genesis
topic Computer Vision and Pattern Recognition
Computation and Language
url https://arxiv.org/abs/2303.11331