[CVPR'24 Highlight] GPT4Point: A Unified Framework for Point-Language Understanding and Generation.
See the codeREADME
[CVPR2024] GPT4Point 
: A Unified Framework for Point-Language Understanding and Generation
π₯ News
π₯ 2024/04/27: We have modified the point encoder section, and now evaluation is more functional, although the training section still needs modification.
π₯ 2024/04/13: We release the GPT4Point v1.0, including training and 3D captioning evluation code.
π₯ 2024/04/05: Our paper GPT4Point is selected as CVPR'24 Highlight 2.84% (324/11532) !
π₯ 2024/02/27: Our paper GPT4Point is accepted by CVPR'24!
π₯ 2024/01/19: We release the Objaverse-XL (Point Cloud Format) Download and Extraction way.
π₯ 2023/12/05: The paper GPT4Point (arxiv) has been released, we unified the Point-language Understanding and Generation.
π₯ 2023/08/13: Two-stage Pre-training code of PointBLIP has been released.
π₯ 2023/08/13: Part of datasets used and result files has been uploaded.
π Overview
This project presents GPT4Point , a 3D multi-modality model that aligns 3D point clouds with language. More details are shown in project page.
-
Unified Framework for Point-language Understanding and Generation. We present the unified framework for point-language understanding and generation GPT4Point, including the 3D MLLM for point-text tasks and controlled 3D generation.
-
Automated Point-language Dataset Annotation Engine Pyramid-XL. We introduce the automated point-language dataset annotation engine Pyramid-XL based on Objaverse-XL, currently encompassing 1M pairs of varying levels of coarseness and can be extended cost-effectively.
-
Object-level Point Cloud Benchmark. Establishing a novel object-level point cloud benchmark with comprehensive evaluation metrics for 3D point cloud language tasks. This benchmark thoroughly assesses models' understanding capabilities and facilitates the evaluation of generated 3D objects.
π§ Version
- v1.0 (2024/04/13). We release the training and evaluation (3D captioning) code.
Dataset and text annotation: Cap3D.
LLM Model: OPT 2.7b
π§ Installation
- (Optional) Creating conda environment
conda create -n gpt4point python=3.8
conda activate gpt4point
- install from PyPI
pip install salesforce-lavis
- Or, for development, you may build from source
git clone https://github.com/salesforce/LAVIS.git
cd LAVIS
pip install -e .
π¦ Data Preparation
-
Annotations: All annotations will be downloaded automaticly through hugging_face.
-
Point Cloud: You can download the Cap3D point cloud dataset through the Google Drive Link. You should unzip these 10 tar.gz files and then put them together. and the all folder strucure is:
GPT4Point
βββ data
β βββ cap3d
β β βββ points
β β β βββ Cap3D_pcs_8192_xyz_w_color
β β β β βββ <point cloud id>.pkl
β β β β βββ ...
β β β β βββ <point cloud id>.pkl
β β βββ annotations
β β β βββ cap3d_caption_train.json
β β β βββ cap3d_caption_val.json
β β β βββ cap3d_real_and_chatgpt_caption_test.json
β β β βββ cap3d_real_and_chatgpt_caption_test_gt.json (for evaluation)
π Training
- For stage 1 training:
python -m torch.distributed.run --master_port=32339 --nproc_per_node=4 train.py --cfg-path lavis/projects/gpt4point/train/pretrain_stage1_cap3d.yaml
- For stage 2 training:
python -m torch.distributed.run --master_port=32339 --nproc_per_node=4 train.py --cfg-path lavis/projects/gpt4point/train/pretrain_stage2_cap3d_opt2.7b.yaml
π Evaluation
python -m torch.distributed.run --master_port=32239 --nproc_per_node=1 evaluate.py --cfg-path lavis/projects/gpt4point/eval/captioning3d_cap3d_opt2.7b_eval.yaml
π¦ Point Dataset and Data Annotation Engine (Optional)
Objaverse-XL Point Dataset Download Way
Note that you should cd in the Objaverse-xl_Download directory.
cd ./Objaverse-xl_Download
Then please see the folder Objaverse-xl_Download for details.
Objaverse-XL Point Cloud Data Generation
Please see the Extract_Pointcloud for details.
π TODO List
Dataset and Data Engine
- [β] Release the arxiv and the project page.
- [β] Release the dataset (Objaverse-Xl) Download way.
- [β] Release the dataset (Objaverse-Xl) rendering (points) way.
- [β] Release pretrain training code and 3D captioning val code.
- Release dataset and data annotation engine (Pyramid-XL).
- Release more evaluation code.
- Release more trainingn code.
- Release more models.
π Citation
If you find our work helpful, please cite:
@inproceedings{GPT4Point,
title={GPT4Point: A Unified Framework for Point-Language Understanding and Generation},
author={Zhangyang Qi and Ye Fang and Zeyi Sun and Xiaoyang Wu and Tong Wu and Jiaqi Wang and Dahua Lin and Hengshuang Zhao},
booktitle={CVPR},
year={2024},
}
π License
This work is under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
π Related Work
Together, Let's make LLM for 3D great!
- Point-Bind & Point-LLM: It aligns point clouds with Image-Bind to reason multi-modality input without 3D-instruction data training.
- 3D-LLM: employs 2D foundation models to encode multi-view images of 3D point clouds.
- PointLLM: employs 3D point clouds with LLaVA.
Similar repos
Languages
Python
99.1%
[CVPR'24 Highlight] GPT4Point: A Unified Framework for Point-Language Understanding and Generation.
See the codeREADME
[CVPR2024] GPT4Point : A Unified Framework for Point-Language Understanding and Generation
π₯ News
π₯ 2024/04/27: We have modified the point encoder section, and now evaluation is more functional, although the training section still needs modification.
π₯ 2024/04/13: We release the GPT4Point v1.0, including training and 3D captioning evluation code.
π₯ 2024/04/05: Our paper GPT4Point is selected as CVPR'24 Highlight 2.84% (324/11532) !
π₯ 2024/02/27: Our paper GPT4Point is accepted by CVPR'24!
π₯ 2024/01/19: We release the Objaverse-XL (Point Cloud Format) Download and Extraction way.
π₯ 2023/12/05: The paper GPT4Point (arxiv) has been released, we unified the Point-language Understanding and Generation.
π₯ 2023/08/13: Two-stage Pre-training code of PointBLIP has been released.
π₯ 2023/08/13: Part of datasets used and result files has been uploaded.
π Overview
This project presents GPT4Point , a 3D multi-modality model that aligns 3D point clouds with language. More details are shown in project page.
-
Unified Framework for Point-language Understanding and Generation. We present the unified framework for point-language understanding and generation GPT4Point, including the 3D MLLM for point-text tasks and controlled 3D generation.
-
Automated Point-language Dataset Annotation Engine Pyramid-XL. We introduce the automated point-language dataset annotation engine Pyramid-XL based on Objaverse-XL, currently encompassing 1M pairs of varying levels of coarseness and can be extended cost-effectively.
-
Object-level Point Cloud Benchmark. Establishing a novel object-level point cloud benchmark with comprehensive evaluation metrics for 3D point cloud language tasks. This benchmark thoroughly assesses models' understanding capabilities and facilitates the evaluation of generated 3D objects.
π§ Version
- v1.0 (2024/04/13). We release the training and evaluation (3D captioning) code.
Dataset and text annotation: Cap3D.
LLM Model: OPT 2.7b
π§ Installation
- (Optional) Creating conda environment
conda create -n gpt4point python=3.8
conda activate gpt4point
- install from PyPI
pip install salesforce-lavis
- Or, for development, you may build from source
git clone https://github.com/salesforce/LAVIS.git
cd LAVIS
pip install -e .
π¦ Data Preparation
-
Annotations: All annotations will be downloaded automaticly through hugging_face.
-
Point Cloud: You can download the Cap3D point cloud dataset through the Google Drive Link. You should unzip these 10 tar.gz files and then put them together. and the all folder strucure is:
GPT4Point
βββ data
β βββ cap3d
β β βββ points
β β β βββ Cap3D_pcs_8192_xyz_w_color
β β β β βββ <point cloud id>.pkl
β β β β βββ ...
β β β β βββ <point cloud id>.pkl
β β βββ annotations
β β β βββ cap3d_caption_train.json
β β β βββ cap3d_caption_val.json
β β β βββ cap3d_real_and_chatgpt_caption_test.json
β β β βββ cap3d_real_and_chatgpt_caption_test_gt.json (for evaluation)
π Training
- For stage 1 training:
python -m torch.distributed.run --master_port=32339 --nproc_per_node=4 train.py --cfg-path lavis/projects/gpt4point/train/pretrain_stage1_cap3d.yaml
- For stage 2 training:
python -m torch.distributed.run --master_port=32339 --nproc_per_node=4 train.py --cfg-path lavis/projects/gpt4point/train/pretrain_stage2_cap3d_opt2.7b.yaml
π Evaluation
python -m torch.distributed.run --master_port=32239 --nproc_per_node=1 evaluate.py --cfg-path lavis/projects/gpt4point/eval/captioning3d_cap3d_opt2.7b_eval.yaml
π¦ Point Dataset and Data Annotation Engine (Optional)
Objaverse-XL Point Dataset Download Way
Note that you should cd in the Objaverse-xl_Download directory.
cd ./Objaverse-xl_Download
Then please see the folder Objaverse-xl_Download for details.
Objaverse-XL Point Cloud Data Generation
Please see the Extract_Pointcloud for details.
π TODO List
Dataset and Data Engine
- [β] Release the arxiv and the project page.
- [β] Release the dataset (Objaverse-Xl) Download way.
- [β] Release the dataset (Objaverse-Xl) rendering (points) way.
- [β] Release pretrain training code and 3D captioning val code.
- Release dataset and data annotation engine (Pyramid-XL).
- Release more evaluation code.
- Release more trainingn code.
- Release more models.
π Citation
If you find our work helpful, please cite:
@inproceedings{GPT4Point,
title={GPT4Point: A Unified Framework for Point-Language Understanding and Generation},
author={Zhangyang Qi and Ye Fang and Zeyi Sun and Xiaoyang Wu and Tong Wu and Jiaqi Wang and Dahua Lin and Hengshuang Zhao},
booktitle={CVPR},
year={2024},
}
π License
This work is under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
π Related Work
Together, Let's make LLM for 3D great!
- Point-Bind & Point-LLM: It aligns point clouds with Image-Bind to reason multi-modality input without 3D-instruction data training.
- 3D-LLM: employs 2D foundation models to encode multi-view images of 3D point clouds.
- PointLLM: employs 3D point clouds with LLaVA.
Similar repos
Languages
Python
99.1%