Skip to main content

Image Captioning

Generate captions from fashion images using LLaVA-NeXT.

Usage​

from tryon.preprocessing.captioning import caption_image
from PIL import Image

image = Image.open("fashion_image.jpg")
question = "Extract outfit details as JSON"

json_data, caption = caption_image(
    image=image,
    question=question,
    json_only=False
)

See Preprocessing Overview for details.