[中文版本]

There are a lot of issues now, our team will check and solve them one by one, please wait patiently.

INTERN-2.5: Multimodal Multitask General Large Model

This repository is an official implementation of the InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions.

Paper | Blog in Chinese | Documents

Introduction

SenseTime and Shanghai AI Laboratory jointly released the multimodal multitask general model "INTERN-2.5" on March 14, 2023. "INTERN-2.5" achieved multiple breakthroughs in multimodal multitask processing, and its excellent cross-modal task processing ability in text and image can provide efficient and accurate perception and understanding capabilities for general scenarios such as autonomous driving.

Overview

Highlights

• 👍 The strongest visual universal backbone model with up to 3 billion parameters
• 🏆 Achieved 90.1% Top1 accuracy in ImageNet, the most accurate among open-source models
• 🏆 Achieved 65.5 mAP on the COCO benchmark dataset for object detection, the only model that exceeded 65.0 mAP

News

• Mar 14, 2023: 🚀 "INTERN-2.5" is released！
• Feb 28, 2023: 🚀 InternImage is accepted to CVPR 2023!
• Nov 18, 2022: 🚀 InternImage-XL merged into BEVFormer v2 achieves state-of-the-art performance of 63.4 NDS on nuScenes Camera Only.
• Nov 10, 2022: 🚀 InternImage-H achieves a new record 65.4 mAP on COCO detection test-dev and 62.9 mIoU on ADE20K, outperforming previous models by a large margin.

Applications

1. Performance on Image Modality Tasks

• On the ImageNet benchmark dataset, "INTERN-2.5" achieved a Top-1 accuracy of 90.1% using only publicly available data for image classification. This is the only model, besides two undisclosed models from Google and Microsoft and additional datasets, to achieve a Top-1 accuracy of over 90.0%. It is also the highest-accuracy open-source model on ImageNet and the largest model in scale in the world.
• On the COCO object detection benchmark dataset, "INTERN-2.5" achieved a mAP of 65.5, making it the only model in the world to surpass 65 mAP.
• "INTERN-2.5" achieved the world's best performance on 16 other important visual benchmark datasets, covering classification, detection, and segmentation tasks.

Classification Task

Image Classification	Scene Classification		Long-Tail Classification
ImageNet	Places365	Places 205	iNaturalist 2018
90.1	61.2	71.7	92.3

Detection Task

Conventional Object Detection				Long-Tail Object Detection			Autonomous Driving Object Detection	Dense Object Detection
COCO	VOC 2007	VOC 2012	OpenImage	LVIS minival	LVIS val	BDD100K	nuScenes	CrowdHuman
65.5	94.0	97.2	74.1	65.8	63.2	38.8	64.8	97.2

Segmentation Task

Semantic Segmentation			Street Segmentation	RGBD Segmentation
ADE20K	COCO Stuff-10K	Pascal Context	CityScapes	NYU Depth V2
62.9	59.6	70.3	86.1	69.7

2. Cross-Modal Performance for Image and Text Tasks

• Image-Text Retrieval

"INTERN-2.5" can quickly locate and retrieve the most semantically relevant images based on textual content requirements. This capability can be applied to both videos and image collections and can be further combined with object detection boxes to enable a variety of applications, helping users quickly and easily find the required image resources. For example, it can return the relevant images specified by the text in the album.

• Image-To-Text

"INTERN-2.5" has a strong understanding capability in various aspects of visual-to-text tasks such as image captioning, visual question answering, visual reasoning, and optical character recognition. For example, in the context of autonomous driving, it can enhance the scene perception and understanding capabilities, assist the vehicle in judging traffic signal status, road signs, and other information, and provide effective perception information support for vehicle decision-making and planning.

Multimodal Tasks

Image Captioning	Fine-tuning Image-Text Retrieval		Zero-shot Image-Text Retrieval
COCO Caption	COCO Caption	Flickr30k	Flickr30k
148.2	76.4	94.8	89.1

Core Technologies

The outstanding performance of "INTERN-2.5" in the field of cross-modal learning is due to several innovations in the core technology of multi-modal multi-task general model, including the development of InternImage as the backbone network for visual perception, LLM as the large-scale text pre-training network for text processing, and Uni-Perceiver as the compatible decoding modeling for multi-task learning.

InternImage, the visual backbone network of "INTERN-2.5", has a parameter size of up to 3 billion and can adaptively adjust the position and combination of convolutions based on dynamic sparse convolution operators, providing powerful representations for multi-functional visual perception. Uni-Perceiver, a versatile task decoding model, encodes data from different modalities into a unified representation space and unifies different tasks into the same task paradigm, enabling simultaneous processing of various modalities and tasks with the same task architecture and shared model parameters.

Project Release

• Model for other downstream tasks
• InternImage-H(1B)/G(3B)
• TensorRT inference
• Classification code of the InternImage series
• InternImage-T/S/B/L/XL ImageNet-1K pretrained model
• InternImage-L/XL ImageNet-22K pretrained model
• InternImage-T/S/B/L/XL detection and instance segmentation model
• InternImage-T/S/B/L/XL semantic segmentation model

Related Projects

• Object Detection and Instance Segmentation: COCO
• Semantic Segmentation: ADE20K, Cityscapes
• Image-Text Retrieval, Image Captioning, and Visual Question Answering: Uni-Perceiver
• 3D Perception: BEVFormer

Open-source Visual Pretrained Models

name	pretrain	pre-training resolution	#param	download
InternImage-L	ImageNet-22K	384x384	223M	ckpt
InternImage-XL	ImageNet-22K	384x384	335M	ckpt
InternImage-H	Joint 427M	384x384	1.08B	ckpt
InternImage-G	-	384x384	3B	ckpt

ImageNet-1K Image Classification

name	pretrain	resolution	acc@1	#param	FLOPs	download
InternImage-T	ImageNet-1K	224x224	83.5	30M	5G	ckpt | cfg
InternImage-S	ImageNet-1K	224x224	84.2	50M	8G	ckpt | cfg
InternImage-B	ImageNet-1K	224x224	84.9	97M	16G	ckpt | cfg
InternImage-L	ImageNet-22K	384x384	87.7	223M	108G	ckpt | cfg
InternImage-XL	ImageNet-22K	384x384	88.0	335M	163G	ckpt | cfg
InternImage-H	Joint 427M	640x640	89.6	1.08B	1478G	ckpt | cfg
InternImage-G	-	512x512	90.1	3B	2700G	ckpt | cfg

COCO Object Detection and Instance Segmentation

backbone	method	schd	box mAP	mask mAP	#param	FLOPs	download
InternImage-T	Mask R-CNN	1x	47.2	42.5	49M	270G	ckpt | cfg
InternImage-T	Mask R-CNN	3x	49.1	43.7	49M	270G	ckpt | cfg
InternImage-S	Mask R-CNN	1x	47.8	43.3	69M	340G	ckpt | cfg
InternImage-S	Mask R-CNN	3x	49.7	44.5	69M	340G	ckpt | cfg
InternImage-B	Mask R-CNN	1x	48.8	44.0	115M	501G	ckpt | cfg
InternImage-B	Mask R-CNN	3x	50.3	44.8	115M	501G	ckpt | cfg
InternImage-L	Cascade	1x	54.9	47.7	277M	1399G	ckpt | cfg
InternImage-L	Cascade	3x	56.1	48.5	277M	1399G	ckpt | cfg
InternImage-XL	Cascade	1x	55.3	48.1	387M	1782G	ckpt | cfg
InternImage-XL	Cascade	3x	56.2	48.8	387M	1782G	ckpt | cfg

backbone	method	box mAP (val/test)	#param	FLOPs	download
InternImage-H	DINO (TTA)	65.0 / 65.4	2.18B	TODO	TODO
InternImage-G	DINO (TTA)	65.3 / 65.5	3B	TODO	TODO

ADE20K Semantic Segmentation

backbone	method	resolution	mIoU (ss/ms)	#param	FLOPs	download
InternImage-T	UperNet	512x512	47.9 / 48.1	59M	944G	ckpt | cfg
InternImage-S	UperNet	512x512	50.1 / 50.9	80M	1017G	ckpt | cfg
InternImage-B	UperNet	512x512	50.8 / 51.3	128M	1185G	ckpt | cfg
InternImage-L	UperNet	640x640	53.9 / 54.1	256M	2526G	ckpt | cfg
InternImage-XL	UperNet	640x640	55.0 / 55.3	368M	3142G	ckpt | cfg
InternImage-H	UperNet	896x896	59.9 / 60.3	1.12B	3566G	ckpt | cfg
InternImage-H	Mask2Former	896x896	62.5 / 62.9	1.31B	4635G	TODO

Main Results of FPS

export classification model from pytorch to tensorrt

export detection model from pytorch to tensorrt

export segmentation model from pytorch to tensorrt

name	resolution	#param	FLOPs	batch 1 FPS (TensorRT)
InternImage-T	224x224	30M	5G	156
InternImage-S	224x224	50M	8G	129
InternImage-B	224x224	97M	16G	116
InternImage-L	384x384	223M	108G	56
InternImage-XL	384x384	335M	163G	47

Before using mmdeploy to convert our PyTorch models to TensorRT, please make sure you have the DCNv3 custom operator builded correctly. You can build it with the following command:

export MMDEPLOY_DIR=/the/root/path/of/MMDeploy

# prepare our custom ops, you can find it at InternImage/tensorrt/modulated_deform_conv_v3
cp -r modulated_deform_conv_v3 ${MMDEPLOY_DIR}/csrc/mmdeploy/backend_ops/tensorrt

# build custom ops
cd ${MMDEPLOY_DIR}
mkdir -p build && cd build
cmake -DCMAKE_CXX_COMPILER=g++-7 -DMMDEPLOY_TARGET_BACKENDS=trt -DTENSORRT_DIR=${TENSORRT_DIR} -DCUDNN_DIR=${CUDNN_DIR} ..
make -j$(nproc) && make install

# install the mmdeploy after building custom ops
cd ${MMDEPLOY_DIR}
pip install -e .

For more details on building custom ops, please refering to this document.

Citation

If this work is helpful for your research, please consider citing the following BibTeX entry.

@article{wang2022internimage,
  title={InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutions},
  author={Wang, Wenhai and Dai, Jifeng and Chen, Zhe and Huang, Zhenhang and Li, Zhiqi and Zhu, Xizhou and Hu, Xiaowei and Lu, Tong and Lu, Lewei and Li, Hongsheng and others},
  journal={arXiv preprint arXiv:2211.05778},
  year={2022}
}

@inproceedings{zhu2022uni,
  title={Uni-perceiver: Pre-training unified architecture for generic perception for zero-shot and few-shot tasks},
  author={Zhu, Xizhou and Zhu, Jinguo and Li, Hao and Wu, Xiaoshi and Li, Hongsheng and Wang, Xiaohua and Dai, Jifeng},
  booktitle={CVPR},
  pages={16804--16815},
  year={2022}
}

@article{zhu2022uni,
  title={Uni-perceiver-moe: Learning sparse generalist models with conditional moes},
  author={Zhu, Jinguo and Zhu, Xizhou and Wang, Wenhai and Wang, Xiaohua and Li, Hongsheng and Wang, Xiaogang and Dai, Jifeng},
  journal={arXiv preprint arXiv:2206.04674},
  year={2022}
}

@article{li2022uni,
  title={Uni-Perceiver v2: A Generalist Model for Large-Scale Vision and Vision-Language Tasks},
  author={Li, Hao and Zhu, Jinguo and Jiang, Xiaohu and Zhu, Xizhou and Li, Hongsheng and Yuan, Chun and Wang, Xiaohua and Qiao, Yu and Wang, Xiaogang and Wang, Wenhai and others},
  journal={arXiv preprint arXiv:2211.09808},
  year={2022}
}

@article{yang2022bevformer,
  title={BEVFormer v2: Adapting Modern Image Backbones to Bird's-Eye-View Recognition via Perspective Supervision},
  author={Yang, Chenyu and Chen, Yuntao and Tian, Hao and Tao, Chenxin and Zhu, Xizhou and Zhang, Zhaoxiang and Huang, Gao and Li, Hongyang and Qiao, Yu and Lu, Lewei and others},
  journal={arXiv preprint arXiv:2211.10439},
  year={2022}
}

@article{su2022towards,
  title={Towards All-in-one Pre-training via Maximizing Multi-modal Mutual Information},
  author={Su, Weijie and Zhu, Xizhou and Tao, Chenxin and Lu, Lewei and Li, Bin and Huang, Gao and Qiao, Yu and Wang, Xiaogang and Zhou, Jie and Dai, Jifeng},
  journal={arXiv preprint arXiv:2211.09807},
  year={2022}
}

@inproceedings{li2022bevformer,
  title={Bevformer: Learning bird’s-eye-view representation from multi-camera images via spatiotemporal transformers},
  author={Li, Zhiqi and Wang, Wenhai and Li, Hongyang and Xie, Enze and Sima, Chonghao and Lu, Tong and Qiao, Yu and Dai, Jifeng},
  booktitle={ECCV},
  pages={1--18},
  year={2022},
}