ijepa_vitg16_22k

How does it work?

As opposed to generative methods that have a pixel decoder, I-JEPA has a predictor that makes predictions in latent space. The predictor in I-JEPA can be seen as a primitive (and restricted) world-model that is able to model spatial uncertainty in a static image from a partially observable context. This world model is semantic in the sense that it predicts high level information about unseen regions in the image, rather than pixel-level details.

We trained a stochastic decoder that maps the I-JEPA predicted representations back in pixel space as sketches. The model correctly captures positional uncertainty and produces high-level object parts with the correct pose (e.g., dog’s head, wolf’s front legs).

Intended uses & limitations

I-JEPA can be used for image classification or feature extraction. This checkpoint in specific is intended for Feature Extraction.

How to use

Here is how to use this model for image feature extraction:

import requests
from PIL import Image
from torch.nn.functional import cosine_similarity

from transformers import AutoModel, AutoProcessor

url_1 = "http://images.cocodataset.org/val2017/000000039769.jpg"
url_2 = "http://images.cocodataset.org/val2017/000000219578.jpg"
image_1 = Image.open(requests.get(url_1, stream=True).raw)
image_2 = Image.open(requests.get(url_2, stream=True).raw)

model_id = "jmtzt/ijepa_vitg16_22k"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModel.from_pretrained(model_id)

def infer(image):
    inputs = processor(image, return_tensors="pt")
    outputs = model(**inputs)
    return outputs.last_hidden_state.mean(dim=1)

embed_1 = infer(image_1)
embed_2 = infer(image_2)

similarity = cosine_similarity(embed_1, embed_2)
print(similarity)

BibTeX entry and citation info

If you use I-JEPA or this code in your work, please cite:

@article{assran2023self,
  title={Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture},
  author={Assran, Mahmoud and Duval, Quentin and Misra, Ishan and Bojanowski, Piotr and Vincent, Pascal and Rabbat, Michael and LeCun, Yann and Ballas, Nicolas},
  journal={arXiv preprint arXiv:2301.08243},
  year={2023}
}