---
license: mit
pipeline_tag: image-segmentation
---

# Retinal Artery-Vein Segmentation Models



## Available checkpoints

We present the OCULARNet model, a base UNet encoder - (256, 128, 64, 32) layer size - coupled with a RepVGG-B3 decoder loaded from `segmentation_models_pytorch` for retinal artery-vein segmentation.

Datasets, training and inference details can be found in our [GitHub repository](`https://github.com/GonzaloPlaaza/OCULAR`).

## Usage

```python
import torch

device = torch.device('cuda:0' if use_cuda else 'cpu')
checkpoint = torch.load("OCULARNet.pth", map_location=device)
        
if "model_state_dict" in checkpoint:
    state_dict = checkpoint["model_state_dict"]
else:
    state_dict = checkpoint

model.load_state_dict(state_dict)