Learning ECG Image Representations via Dual Physiological-Aware Alignments

Quickstart

from transformers import AutoModel, CLIPImageProcessor
from PIL import Image
import torch

model = AutoModel.from_pretrained("Manhph2211/ECG-Scan", trust_remote_code=True)
model.eval()

processor = CLIPImageProcessor.from_pretrained("openai/clip-vit-large-patch14-336")
img = Image.open("ecg.png").convert("RGB")
pixel_values = processor(images=img, return_tensors="pt")["pixel_values"]

with torch.no_grad():
    out = model(pixel_values).embeddings         

Citation

@article{pham2026learning,
  title={Learning ECG Image Representations via Dual Physiological-Aware Alignments},
  author={Pham, Hung Manh and Tang, Jialu and Saeed, Aaqib and Ma, Dong and Zhu, Bin and Zhou, Pan},
  journal={arXiv preprint arXiv:2604.01526},
  year={2026}
}