SpatialGen

Image-to-Scene Results	Text-to-Scene Results

SpatialGen produces multi-view, multi-modal information from a semantic layout using a multi-view, multi-modal diffusion model.

• [Aug, 2025] Initial release of SpatialGen-1.0!

• Provide inference code of SpatialGen.
• Provide training instruction for SpatialGen.
• Release SpatialGen dataset.

Tested with the following environment:

• Python 3.10
• PyTorch 2.3.1
• CUDA Version 12.1

# clone the repository
git clone https://github.com/manycore-research/SpatialGen.git
cd SpatialGen

python -m venv .venv
source .venv/bin/activate

pip install -r requirements.txt
# Optional: fix the [flux inference bug](https://github.com/vllm-project/vllm/issues/4392)
pip install nvidia-cublas-cu12==12.4.5.8

We provide SpatialGen-Testset with 48 rooms, which labeled with 3D layout and 4.8K rendered images (48 x 100 views, including RGB, normal, depth maps and semantic maps) for MVD inference.

# Single image-to-3D Scene
bash scripts/infer_spatialgen_i2s.sh

# Text-to-image-to-3D Scene
bash scripts/infer_spatialgen_t2s.sh

SpatialGen-1.0 is derived from Stable-Diffusion-v2.1, which is licensed under the CreativeML Open RAIL++-M License. FLUX.1-Layout-ControlNet is licensed under the FLUX.1-dev Non-Commercial License.

We would like to thank the following projects that made this work possible:

DiffSplat | SD 2.1 | TAESD | FLUX | SpatialLM