license: cc-by-4.0
dataset_info:
features:
- name: img_id
dtype: string
- name: turn_index
dtype: int32
- name: input_img
dtype: image
- name: instruction
dtype: string
- name: output_img
dtype: image
splits:
- name: train
num_bytes: 15474910323.726
num_examples: 8807
- name: dev
num_bytes: 926910159
num_examples: 528
download_size: 16474461666
dataset_size: 16401820482.726
task_categories:
- text-to-image
language:
- en
pretty_name: MagicBrush
size_categories:
- 10K<n<100K
Dataset Card for Dataset Name
Dataset Summary
MagicBrush is the first large-scale, manually-annotated instruction-guided image editing dataset covering diverse scenarios single-turn, multi-turn, mask-provided, and mask-free editing. MagicBrush comprises 10K (source image, instruction, target image) triples, which is sufficient to train large-scale image editing models.
Please check our website to explore more visual results.
Dataset Structure
"img_id" (str): same from COCO id "turn_index" (int32): the edit turn in the image "input_img" (str): input image, could be the original real image (turn_index=1) and edited images from last tunr (turn_index >=2) "instruction" (str): edit instruction of how the input image should be changed. "output_img" (str): the edited image corresponding to the input image and instruction.
If you need auxiliary data, please use training set and dev set
Splits
train: 8,807 edit turns (4,512 edit sessions). val: 528 edit turns (266 edit sessions). test: (To prevent potential data leakage, please check our repo for information on obtaining the test set.)
Licensing Information
Creative Commons License This work is licensed under a Creative Commons Attribution 4.0 International License.
Citation Information
TODO