Code

Open code & datasets.

Reference implementations on GitHub. Datasets are released as licensing and review allow.

Public code lives on the PediaMedAI GitHub organization (opens in new tab) . Reference implementations are released alongside our published research; clinical-data licensing constraints determine what we can release and when.

— Repositories

Open implementations.

2022 · IJCAIOpen dataset

AggPose

Deep aggregation vision transformer for infant pose estimation. Cao et al., 2022.

★ 31PythonUpdated
github.com/PediaMedAI/AggPose
(opens in new tab)
2023 · ICASSPASD diagnosis

ViTASD

Robust vision transformer baselines for autism spectrum disorder facial diagnosis. Cao et al., 2023.

★ 29PythonUpdated
github.com/PediaMedAI/ViTASD
(opens in new tab)
2023 · UAICalibration

LRFormer

Mitigating transformer overconfidence via Lipschitz regularization. Ye et al., 2023.

★ 0PythonUpdated
github.com/IrohXu/Lipschitz-Transformer-PyTorch
(opens in new tab)
PreprintDisease progression

PIE

Simulating disease progression via progressive image editing. Liang et al., 2023.

★ 30PythonUpdated
github.com/PediaMedAI/PIE
(opens in new tab)