[ICASSP 2024] SGM: A Dataset for 3D Garment Reconstruction from Single Hand-Drawn Sketch

News

• Now, SGM encompasses a total of 15 garment categories, comprising 1,471 instances of garment 3D models.
• Additionally, depth maps have been added to the 2D renderings of each garment.
• The download link for SGM: Google Drive: Google Drive

---

Overviews

We propose SGM, a comprehensive dataset comprising 656 garment models categorized into short and long sleeves. Each garment model includes two types of annotations:

• Renderings from 12 perspectives: sketches, contours, silhouettes, and normal maps, resulting in a total of 48 renderings per garment model.
• Spatial sampling points and UDF (Unsigned Distance Function) values from each sampling point to the surface of the garment model.

The dataset is currently being expanded and produced, and we will make it publicly available after the publication of our paper.

---

Citation

If you use SGM in your work, please consider citing our paper!

@INPROCEEDINGS{10447299,
  author={Chen, Jia and Qin, Jinlong and Zhong, Saishang and Yang, Kai and Hu, Xinrong and Peng, Tao and Li, Rui},
  booktitle={ICASSP 2024 - 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)}, 
  title={SGM: A Dataset for 3D Garment Reconstruction from Single Hand-Drawn Sketch}, 
  year={2024},
  pages={2510-2514},
  keywords={Solid modeling;Three-dimensional displays;Clothing;Signal processing;Reconstruction algorithms;Acoustics;Speech processing;Garment reconstruction;Sketch;UDF;Dataset;Deep generative models},
  doi={10.1109/ICASSP48485.2024.10447299}
}