Depth Estimation
Core ML
Depth Pro
visionos
apple-silicon
amlr
computer-vision
512x512
ane-optimized
Instructions to use aarondevstack/DepthPro-512x512-coreml with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Depth Pro
How to use aarondevstack/DepthPro-512x512-coreml with Depth Pro:
# Download checkpoint pip install huggingface-hub huggingface-cli download --local-dir checkpoints aarondevstack/DepthPro-512x512-coreml
import depth_pro # Load model and preprocessing transform model, transform = depth_pro.create_model_and_transforms() model.eval() # Load and preprocess an image. image, _, f_px = depth_pro.load_rgb("example.png") image = transform(image) # Run inference. prediction = model.infer(image, f_px=f_px) # Results: 1. Depth in meters depth = prediction["depth"] # Results: 2. Focal length in pixels focallength_px = prediction["focallength_px"] - Notebooks
- Google Colab
- Kaggle
metadata
license: apple-ascl
library_name: coreml
tags:
- apple
- amlr
- computer-vision
- depth-estimation
- visionos
- macos
DepthPro: High-Resolution Monocular Depth (CoreML)
This repository provides CoreML-optimized weights for DepthPro, adapted for on-device inference on visionOS and macOS.
Configuration
- Resolution: 512x512 (Optimized for <1s inference)
- Precision: Float16
- Architecture: Multi-scale ViT-Large with Sharp Boundary Refinement
Performance (M2 Apple Vision Pro)
- Transform + Encoder: ~700ms (GPU Fallback)
- Decoder + Depth Head: ~150ms
- Total Pipeline: < 1.0s per frame
Citation
If you use this model in your research or applications, please cite the original work:
@article{Bochkovskii2024:DepthPro,
title={DepthPro: Sharp Monocular Depth Estimation with Full-Resolution Refinement},
author={Bochkovskii, Alexey and Fogosar, Ambron and Koltun, Vladlen},
journal={arXiv preprint arXiv:2410.02073},
year={2024}
}