This paper presents a grounded language-image pre-training (GLIP) model for learning object-level, language-aware, and semantic-rich visual representations. GLIP unifies object detection and phrase grounding for pre-training. The unification brings two benefits: 1) it allows GLIP to learn from both detection and grounding data to improve both tasks and bootstrap a good grounding model; 2) GLIP can leverage massive image-text pairs by generating grounding boxes in a self-training fashion, making the learned representation semantic-rich. In our experiments, we pre-train GLIP on 27M grounding data, including 3M human-annotated and 24M web-crawled image-text pairs. The learned representations demonstrate strong zero-shot and few-shot transferability to various object-level recognition tasks. 1) When directly evaluated on COCO and LVIS (without seeing any images in COCO during pre-training), GLIP achieves 49.8 AP and 26.9 AP, respectively, surpassing many supervised baselines. 2) After fine-tuned on COCO, GLIP achieves 60.8 AP on val and 61.5 AP on test-dev, surpassing prior SoTA. 3) When transferred to 13 downstream object detection tasks, a 1-shot GLIP rivals with a fully-supervised Dynamic Head.
Use the model
import torch
from mmpretrain import get_model
model = get_model('swin-t_glip-pre_3rdparty', pretrained=True)
inputs = torch.rand(1, 3, 224, 224)
out = model(inputs)
print(type(out))
# To extract features.
feats = model.extract_feat(inputs)
print(type(feats))
The pre-trained models are used to fine-tune, and therefore don't have evaluation results.
Model | Pretrain | resolution | Download |
---|---|---|---|
GLIP-T (swin-t_glip-pre_3rdparty )* |
O365,GoldG,CC3M,SBU | 224x224 | model |
GLIP-L (swin-l_glip-pre_3rdparty_384px )* |
FourODs,GoldG,CC3M+12M,SBU | 384x384 | model |
Models with * are converted from the official repo.
@inproceedings{li2021grounded,
title={Grounded Language-Image Pre-training},
author={Liunian Harold Li* and Pengchuan Zhang* and Haotian Zhang* and Jianwei Yang and Chunyuan Li and Yiwu Zhong and Lijuan Wang and Lu Yuan and Lei Zhang and Jenq-Neng Hwang and Kai-Wei Chang and Jianfeng Gao},
year={2022},
booktitle={CVPR},
}