| --- |
| license: other |
| license_name: apple-ml-research-model-license |
| license_link: https://github.com/apple/ml-sharp/blob/main/LICENSE_MODEL |
| library_name: coreml |
| tags: |
| - coreml |
| - ios |
| - view-synthesis |
| - 3d-gaussian-splatting |
| - monocular-view-synthesis |
| - openreshoot |
| - apple-sharp |
| --- |
| |
| # OpenReshoot SHARP Core ML Package |
|
|
| This repository hosts the Core ML package used by OpenReshoot for on-device |
| single-image view synthesis experiments on iOS. |
|
|
| The model is a Core ML conversion of Apple's SHARP research checkpoint. SHARP |
| predicts a 3D Gaussian representation from one input image, enabling nearby |
| view rendering from a single photograph. |
|
|
| ## Files |
|
|
| ```text |
| SHARP.mlpackage/ |
| Manifest.json |
| Data/com.apple.CoreML/model.mlmodel |
| Data/com.apple.CoreML/weights/weight.bin |
| ``` |
|
|
| OpenReshoot downloads these files and reconstructs the `SHARP.mlpackage` |
| directory locally. |
|
|
| ## Intended Use |
|
|
| - Research and experimentation with monocular view synthesis. |
| - OpenReshoot iOS builds that download the model instead of bundling it inside |
| the app. |
| - Local Core ML testing of the converted SHARP model. |
|
|
| ## Not Intended For |
|
|
| - Commercial products or services. |
| - General-purpose photo editing, identity editing, surveillance, or biometric |
| use. |
| - Text-to-image generation or image restoration. This package is for SHARP's |
| single-image 3D representation prediction path. |
|
|
| ## Source And Conversion |
|
|
| Upstream project: |
|
|
| - Apple ML SHARP: https://github.com/apple/ml-sharp |
| - Project page: https://apple.github.io/ml-sharp/ |
| - Paper: https://arxiv.org/abs/2512.10685 |
|
|
| The upstream checkpoint is converted to Core ML for OpenReshoot using the |
| conversion tooling in the OpenReshoot repository. |
|
|
| ## License |
|
|
| The model weights are covered by the Apple Machine Learning Research Model |
| License Agreement. See `LICENSE_MODEL` in this repository. |
|
|
| Required attribution notice: |
|
|
| > Apple Machine Learning Research Model is licensed under the Apple Machine |
| > Learning Research Model License Agreement. |
|
|
| The license is for research purposes. Review the full license before using, |
| redistributing, or modifying the model. |
|
|
| ## Citation |
|
|
| ```bibtex |
| @inproceedings{Sharp2025:arxiv, |
| title = {Sharp Monocular View Synthesis in Less Than a Second}, |
| author = {Mescheder, Lars and Dong, Wei and Li, Shiwei and Bai, Xuyang and Santos, Marcel and Hu, Peiyun and Lecouat, Bruno and Zhen, Mingmin and Delaunoy, Amaël and Fang, Tian and Tsin, Yanghai and Richter, Stephan and Koltun, Vladlen}, |
| booktitle = {arXiv}, |
| year = {2025} |
| } |
| ``` |
|
|
| ## Notes For OpenReshoot |
|
|
| Use the Hugging Face raw file URLs for these three files when wiring the iOS |
| downloader: |
|
|
| ```text |
| https://huggingface.co/kellyxiaowei/openreshoot-sharp-coreml/resolve/main/SHARP.mlpackage/Manifest.json |
| https://huggingface.co/kellyxiaowei/openreshoot-sharp-coreml/resolve/main/SHARP.mlpackage/Data/com.apple.CoreML/model.mlmodel |
| https://huggingface.co/kellyxiaowei/openreshoot-sharp-coreml/resolve/main/SHARP.mlpackage/Data/com.apple.CoreML/weights/weight.bin |
| ``` |
|
|
