Hongsheng Li 1,2 Yu Qiao 2 Wanli Ouyang2 Xiangyu Yue1,†,‡
2 OpenGVLab,Shanghai AI Laboratory
* Equal Contribution † Corresponding Author ‡ Project Lead
As a foundation model, Meta-Transformer can handle data from 12 modalities, which determines that it can support a wide range of applications. As shown in this figure, Meta-Transformer can provide services for downstream tasks including stock analysis 📈, weather forecasting ☀️ ☔ ☁️ ❄️ ⛄ ⚡, remote sensing 📡, autonomous driving 🚗, social network 🌍, speech recognition 🔉, etc.
Table 1: Meta-Transformer is capable of handling up to 12 modalities, including natural language , RGB images
, point clouds
, audios
, videos
, tabular data
, graph
, time series data
, hyper-spectral images
, IMU
, medical images
, and infrared images
.
This repository is built to explore the potential and extensibility of transformers for multimodal learning. We utilize the advantages of Transformers to deal with length-variant sequences. Then we propose the Data-to-Sequence tokenization following a meta-scheme, then we apply it to 12 modalities including text, image, point cloud, audio, video, infrared, hyper-spectral, X-Ray, tabular, graph, time-series, and Inertial Measurement Unit (IMU) data.
After obtaining the token sequence, we employ a modality-shared encoder to extract representation across different modalities. With task-specific heads, Meta-Transformer can handle various tasks on the different modalities, such as: classification, detection, and segmentation.
- 2023.8.2: 🎉🎉🎉 The implementation of Meta-Transformer for image, point cloud, graph, tabular, time-series, X-Ray, hyper-spectrum, LiDAR data has been released. We also release a very powerful foundation model for Autonomous Driving 🚀🚀🚀.
- 2023.7.22: 🌟🌟🌟 Pretrained weights and a usage demo for our Meta-Transformer have been released. Comprehensive documentation and implementation of the image modality are underway and will be released soon. Stay tuned for more exciting updates!⌛⌛⌛
- 2023.7.21: Paper is released at arxiv, and code will be gradually released.
- 2023.7.8: Github Repository Initialization.
Model | Pretraining | Scale | #Param | Download | 国内下载源 |
---|---|---|---|---|---|
Meta-Transformer-B16 | LAION-2B | Base | 85M | ckpt | ckpt |
Meta-Transformer-L14 | LAION-2B | Large | 302M | ckpt | ckpt |
from timm.models.vision_transformer import Block
ckpt = torch.load("Meta-Transformer_base_patch16_encoder.pth")
encoder = nn.Sequential(*[
Block(
dim=768,
num_heads=12,
mlp_ratio=4.,
qkv_bias=True,
norm_layer=nn.LayerNorm,
act_layer=nn.GELU
)
for i in range(12)])
encoder.load_state_dict(ckpt,strict=True)
- Meta-Transformer with Large Language Models.
- Multimodal Joint Training with Meta-Transformer.
- Support More Modalities and More Tasks.
🚀🚀🚀 We aspire to shape this repository into a formidable foundation for mainstream AI perception tasks across diverse modalities. Your contributions can play a significant role in this endeavor, and we warmly welcome your participation in our project!
To contact us, never hestitate to send an email to [email protected]
,[email protected]
, [email protected]
, or [email protected]
!
If the code and paper help your research, please kindly cite:
@article{zhang2023metatransformer,
title={Meta-Transformer: A Unified Framework for Multimodal Learning},
author={Zhang, Yiyuan and Gong, Kaixiong and Zhang, Kaipeng and Li, Hongsheng and Qiao, Yu and Ouyang, Wanli and Yue, Xiangyu},
year={2023},
journal={arXiv preprint arXiv:2307.10802},
}
This project is released under the Apache 2.0 license.
This code is developed based on excellent open-sourced projects including MMClassification, MMDetection, MMsegmentation, OpenPoints, Time-Series-Library, Graphomer, SpectralFormer, and ViT-Adapter.