Paper abstract:
In this work, we present a conceptually simple and effective method to train a strong bilingual/multilingual multimodal representation model. Starting from the pre-trained multimodal representation model CLIP released by OpenAI, we altered its text encoder with a pre-trained multilingual text encoder XLM-R, and aligned both languages and image representations by a two-stage training schema consisting of teacher learning and contrastive learning. We validate our method through evaluations of a wide range of tasks. We set new state-of-the-art performances on a bunch of tasks including ImageNet-CN, Flicker30k-CN, COCO-CN and XTD. Further, we obtain very close performances with CLIP on almost all tasks, suggesting that one can simply alter the text encoder in CLIP for extended capabilities such as multilingual understanding. Our models and code are available at this https URL.
You can use AltCLIP to calculate image embeddings. These embeddings can be used for:
First, install Autodistill and Autodistill AltCLIP:
pip install autodistill autodistill-altclip
Then, run:
from autodistill_altclip import AltCLIP
from autodistill.detection import CaptionOntology
# define an ontology to map class names to our AltCLIP prompt
# the ontology dictionary has the format {caption: class}
# where caption is the prompt sent to the base model, and class is the label that will
# be saved for that caption in the generated results
# then, load the model
base_model = AltCLIP(
ontology=CaptionOntology(
{
"person": "person",
"a forklift": "forklift"
}
)
)
results = base_model.predict("construction.jpg")
print(results)