Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/cure-lab/magicdrive
[ICLR24] Official implementation of the paper “MagicDrive: Street View Generation with Diverse 3D Geometry Control”
https://github.com/cure-lab/magicdrive
autonomous-vehicles deep-learning diffusion-models image-generation pytorch video-generation
Last synced: 5 days ago
JSON representation
[ICLR24] Official implementation of the paper “MagicDrive: Street View Generation with Diverse 3D Geometry Control”
- Host: GitHub
- URL: https://github.com/cure-lab/magicdrive
- Owner: cure-lab
- License: apache-2.0
- Created: 2023-10-11T08:33:23.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-12-09T09:58:54.000Z (about 2 months ago)
- Last Synced: 2025-01-11T09:02:32.770Z (12 days ago)
- Topics: autonomous-vehicles, deep-learning, diffusion-models, image-generation, pytorch, video-generation
- Language: Python
- Homepage: https://gaoruiyuan.com/magicdrive/
- Size: 23.5 MB
- Stars: 797
- Watchers: 14
- Forks: 45
- Open Issues: 18
-
Metadata Files:
- Readme: README.MD
- License: LICENSE
Awesome Lists containing this project
README
# MagicDrive
✨ Check out our new work [MagicDrive3D](https://github.com/flymin/MagicDrive3D) on **3D scene generation**!
✨ If you want **video generation**, please find the code at the [`video branch`](https://github.com/cure-lab/MagicDrive/tree/video).
[](https://arxiv.org/abs/2310.02601) [](https://gaoruiyuan.com/magicdrive/) [](https://github.com/cure-lab/MagicDrive/blob/main/LICENSE) [](https://github.com/cure-lab/MagicDrive)
Videos generated by MagicDrive (click the image to see the video).
[](https://gaoruiyuan.com/magicdrive/static/videos/2_7_gen.mp4)
[](https://gaoruiyuan.com/magicdrive/static/videos/3_7_gen.mp4)
This repository contains the implementation of the paper
> MagicDrive: Street View Generation with Diverse 3D Geometry Control
> [Ruiyuan Gao](https://gaoruiyuan.com/)1\*, [Kai Chen](https://kaichen1998.github.io/)2\*, [Enze Xie](https://xieenze.github.io/)3^, [Lanqing Hong](https://scholar.google.com.sg/citations?user=2p7x6OUAAAAJ&hl=en)3, [Zhenguo Li](https://scholar.google.com/citations?user=XboZC1AAAAAJ&hl=en)3, [Dit-Yan Yeung](https://sites.google.com/view/dyyeung)2, [Qiang Xu](https://cure-lab.github.io/)1^
> 1CUHK 2HKUST 3Huawei Noah's Ark Lab
> \*Equal Contribution ^Corresponding Authors
## Abstract
TL; DR MagicDrive generates high-quality street-view images & videos with diverse 3D geometry control and multiview consistency, which can be used as a data engine in various perception tasks.
Recent advancements in diffusion models have significantly enhanced the data synthesis with 2D control. Yet, precise 3D control in street view generation, crucial for 3D perception tasks, remains elusive. Specifically, utilizing Bird's-Eye View (BEV) as the primary condition often leads to challenges in geometry control (e.g., height), affecting the representation of object shapes, occlusion patterns, and road surface elevations, all of which are essential to perception data synthesis, especially for 3D object detection tasks. In this paper, we introduce MagicDrive, a novel street view generation framework, offering diverse 3D geometry controls including camera poses, road maps, and 3D bounding boxes, together with textual descriptions, achieved through tailored encoding strategies. Besides, our design incorporates a cross-view attention module, ensuring consistency across multiple camera views. With MagicDrive, we achieve high-fidelity street-view image & video synthesis that captures nuanced 3D geometry and various scene descriptions, enhancing tasks like BEV segmentation and 3D object detection.
## News
- [2024/12/09] We release 60-frame video generation model on [huggingface](https://huggingface.co/flymin/MagicDrive-t-60f-224x400-80k), please use the code in `video` branch to run.
- [2024/12/09] We release two higher-resolution image generation models ([424x800 model](https://huggingface.co/flymin/MagicDrive-424x800-450ep) for visualization and [272x736 model](https://huggingface.co/flymin/MagicDrive-272x736-400ep) for BEVFusion) with their training configs.
- [2024/06/07] MagicDrive can generate **60-frame** videos! We release the config: [rawbox_mv2.0t_0.4.3_60.yaml](https://github.com/cure-lab/MagicDrive/blob/video/configs/exp/rawbox_mv2.0t_0.4.3_60.yaml). Check out our demos on the [project page](https://gaoruiyuan.com/magicdrive/#long-video).
- [2024/06/07] We release **model checkpoint** for **16-frame** video generation. [Check it out](https://github.com/cure-lab/MagicDrive/tree/video?tab=readme-ov-file#magicdrive-t-checkpoints)!
- [2024/06/01] We hold the [W-CODA](https://coda-dataset.github.io/w-coda2024/index.html) workshop @ECCV2024. Challenge [track 2](https://coda-dataset.github.io/w-coda2024/track2/) will use MagicDrive as the baseline. We will release more resources in the near future. Stay tuned!
## Method
In MagicDrive, we employ two strategies (cross-attention and additive encoder branch) to inject text prompts, camera poses, object boxes, and road maps as conditions for generation. We also propose a cross-view attention module for multiview consistency.
## TODO
- [x] [config](configs/exp/224x400.yaml) and [model checkpoint](https://mycuhk-my.sharepoint.com/:u:/g/personal/1155157018_link_cuhk_edu_hk/ERiu-lbAvq5IkODTscFXYPUBpVYVDbwjHchDExBlPfeQ0w?e=8YaDM0) for base resolution (224x400)
- [x] demo for base resolution (224x400)
- [x] GUI for interactive bbox editing
- [x] train and test code release
- [x] FID test code
- [x] config and checkpoint for high resolution
## Getting Started
### Environment Setup
Clone this repo with submodules
```bash
git clone --recursive https://github.com/cure-lab/MagicDrive.git
```
The code is tested with `Pytorch==1.10.2` and `cuda 10.2` on V100 servers. To setup the python environment, follow:
```bash
# option1: to run GUI only
pip install -r requirements/gui.txt
# 😍 our GUI does not need mm-series packages.
# continue to install diffusers from `third_party`.
# option2: to run the full testing demo (and also test your env before training)
cd ${ROOT}
pip install -r requirements/dev.txt
# continue to install `third_party`s as following.
```
We opt to install the source code for the following packages, with `cd ${FOLDER}; pip -vvv install .`
```bash
# install third-party
third_party/
├── bevfusion -> based on db75150
├── diffusers -> based on v0.17.1 (afcca39)
└── xformers -> based on v0.0.19 (8bf59c9), optional
```
see [note about our xformers](doc/xformers.md). If you have issues with the environment setup, please check [FAQ](doc/FAQ.md) first.
Setup default configuration for `accelerate` with
```bash
accelerate config
```
Our default log directory is `${ROOT}/magicdrive-log`. Please be prepared.
### Pretrained Weights
Our training is based on [stable-diffusion-v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5). We assume you put them at `${ROOT}/pretrained/` as follows:
```bash
{ROOT}/pretrained/stable-diffusion-v1-5/
├── text_encoder
├── tokenizer
├── unet
├── vae
└── ...
```
## Street-view Generation with MagicDrive
Download our model checkpoint for MagicDrive from
- 224x400 model: [onedrive](https://mycuhk-my.sharepoint.com/:u:/g/personal/1155157018_link_cuhk_edu_hk/ERiu-lbAvq5IkODTscFXYPUBpVYVDbwjHchDExBlPfeQ0w?e=8YaDM0)
- 272x736 model: [huggingface](https://huggingface.co/flymin/MagicDrive-272x736-400ep)
- 424x800 model: [huggingface](https://huggingface.co/flymin/MagicDrive-424x800-450ep)
and put them in `${ROOT}/pretrained/`
**Run our demo**
👍 We recommend users run our interactive GUI first, because we have minimized the dependencies for the GUI demo.
```bash
cd ${ROOT}
python demo/interactive_gui.py
# a gradio-based gui, use your web browser
```
As suggested by [#37](https://github.com/cure-lab/MagicDrive/issues/37), prompt is configurable through GUI!
Run our demo for camera view generation.
```bash
cd ${ROOT}
python demo/run.py resume_from_checkpoint=magicdrive-log/SDv1.5mv-rawbox_2023-09-07_18-39_224x400
```
The generated images will be located in `magicdrive-log/test`. More information can be find in [demo doc](demo/readme.md).
## Train MagicDrive
### Prepare Data
We prepare the nuScenes dataset similar to [bevfusion's instructions](https://github.com/mit-han-lab/bevfusion#data-preparation). Specifically,
1. Download the nuScenes dataset from the [website](https://www.nuscenes.org/nuscenes) and put them in `./data/`. You should have these files:
```bash
data/nuscenes
├── maps
├── mini
├── samples
├── sweeps
├── v1.0-mini
└── v1.0-trainval
```
> [!TIP]
> You can download the `.pkl` files from [OneDrive](https://mycuhk-my.sharepoint.com/:u:/g/personal/1155157018_link_cuhk_edu_hk/EYF9ZkMHwVZKjrU5CUUPbfYBhC1iZMMnhE2uI2q5iCuv9w?e=QgEmcH). They should be enough for training and testing.
2. Generate mmdet3d annotation files by:
```bash
python tools/create_data.py nuscenes --root-path ./data/nuscenes \
--out-dir ./data/nuscenes_mmdet3d_2 --extra-tag nuscenes
```
You should have these files:
```bash
data/nuscenes_mmdet3d_2
├── nuscenes_dbinfos_train.pkl (-> ${bevfusion-version}/nuscenes_dbinfos_train.pkl)
├── nuscenes_gt_database (-> ${bevfusion-version}/nuscenes_gt_database)
├── nuscenes_infos_train.pkl
└── nuscenes_infos_val.pkl
```
Note: As shown above, some files can be soft-linked with the original version from bevfusion. If some of the files is located in `data/nuscenes`, you can move them to `data/nuscenes_mmdet3d_2` manually.
3. (Optional) To accelerate data loading, we prepared cache files in h5 format for BEV maps. They can be generated through `tools/prepare_map_aux.py` with different configs in `configs/dataset`. For example:
```bash
python tools/prepare_map_aux.py +process=train
python tools/prepare_map_aux.py +process=val
```
You will have files like `./val_tmp.h5` and `./train_tmp.h5`. You have to rename the cache files correctly after generating them. Our default is
```bash
data/nuscenes_map_aux
├── train_26x200x200_map_aux_full.h5 (42G)
└── val_26x200x200_map_aux_full.h5 (9G)
```
### Train the model
Launch training with (with 8xV100):
```bash
accelerate launch --mixed_precision fp16 --gpu_ids all --num_processes 8 tools/train.py \
+exp=224x400 runner=8gpus
```
During training, you can check tensorboard for the log and intermediate results.
Besides, we provide debug config to test your environment and data loading process (with 2xV100):
```bash
accelerate launch --mixed_precision fp16 --gpu_ids all --num_processes 2 tools/train.py \
+exp=224x400 runner=debug runner.validation_before_run=true
```
### Test the model
After training, you can test your model for driving view generation through:
```bash
python tools/test.py resume_from_checkpoint=${YOUR MODEL}
# take our the 224x400 model checkpoint as an example
python tools/test.py resume_from_checkpoint=./pretrained/SDv1.5mv-rawbox_2023-09-07_18-39_224x400
```
Please find the results in `./magicdrive-log/test/`.
**To test FID**
First, you should generate the full validation set with
```bash
python perception/data_prepare/val_set_gen.py \
resume_from_checkpoint=./pretrained/SDv1.5mv-rawbox_2023-09-07_18-39_224x400 \
task_id=224x400 fid.img_gen_dir=./tmp/224x400 +fid=data_gen +exp=224x400
# for map=zero as the null condition for CFG, add `runner.pipeline_param.use_zero_map_as_unconditional=true`
```
For this script, **multi-process / multi-node** is also available by `accelerate`. Just launch it with commands similar to that of training.
Then, test the FID score with
```bash
# we assume your torch cache dir is at "../pretrained/torch_cache/". If you want
# to use the default place, please comment the second last line in "tools/fid_score.py".
python tools/fid_score.py cfg \
resume_from_checkpoint=./pretrained/SDv1.5mv-rawbox_2023-09-07_18-39_224x400 \
fid.rootb=tmp/224x400
```
Alternatively, we provide the pre-generated samples for validation set [here](https://mycuhk-my.sharepoint.com/:f:/g/personal/1155157018_link_cuhk_edu_hk/EjWsTYfC01BAl0F2NLP_bX4BqHjY-oV1VaTx4RgMzbiXWQ?e=fPfEy3).
You can put them in `./tmp` and launch the test through
```bash
python tools/fid_score.py cfg \
resume_from_checkpoint=./pretrained/SDv1.5mv-rawbox_2023-09-07_18-39_224x400 \
fid.rootb=tmp/224x400/samples # FID=14.46065995481922
# or `fid.rootb=tmp/224x400map0/samples`, FID=16.195992872931697
```
## Quantitative Results
Compare MagicDrive with other methods for generation quality:
Training support with images generated from MagicDrive:
More results can be found in the main paper.
## Qualitative Results
More results can be found in the main paper.
## Cite Us
```bibtex
@inproceedings{gao2023magicdrive,
title={{MagicDrive}: Street View Generation with Diverse 3D Geometry Control},
author={Gao, Ruiyuan and Chen, Kai and Xie, Enze and Hong, Lanqing and Li, Zhenguo and Yeung, Dit-Yan and Xu, Qiang},
booktitle = {International Conference on Learning Representations},
year={2024}
}
```
## Credit
We adopt the following open-sourced projects:
- [bevfusion](https://github.com/mit-han-lab/bevfusion): dataloader to handle 3d bounding boxes and BEV map
- [diffusers](https://github.com/huggingface/diffusers): framework to train stable diffusion
- [xformers](https://github.com/facebookresearch/xformers): accelerator for attention mechanism
- Thanks [@pixeli99](https://github.com/pixeli99) for training the [60-frame video generation](https://gaoruiyuan.com/magicdrive/#long-video).