|
|
--- |
|
|
license: cc-by-4.0 |
|
|
dataset_info: |
|
|
features: |
|
|
- name: COCO |
|
|
dtype: image |
|
|
- name: T_original |
|
|
dtype: string |
|
|
- name: T_generated |
|
|
dtype: string |
|
|
- name: T_negative |
|
|
dtype: string |
|
|
- name: NeIn |
|
|
dtype: image |
|
|
splits: |
|
|
- name: train |
|
|
num_bytes: 71681524914.3 |
|
|
num_examples: 342775 |
|
|
- name: validation |
|
|
num_bytes: 4945769234.918 |
|
|
num_examples: 24182 |
|
|
download_size: 36634351590 |
|
|
dataset_size: 76627294149.218 |
|
|
configs: |
|
|
- config_name: default |
|
|
data_files: |
|
|
- split: train |
|
|
path: data/train-* |
|
|
- split: validation |
|
|
path: data/validation-* |
|
|
task_categories: |
|
|
- text-to-image |
|
|
- image-to-image |
|
|
language: |
|
|
- en |
|
|
tags: |
|
|
- image-editing |
|
|
- negative-prompt |
|
|
- diffusion |
|
|
- negation |
|
|
- vlm |
|
|
pretty_name: NeIn |
|
|
size_categories: |
|
|
- 100K<n<1M |
|
|
--- |
|
|
|
|
|
# NeIn: Telling What You Don’t Want (CVPRW 2025) |
|
|
|
|
|
### Dataset Description |
|
|
- **Homepage:** https://tanbuinhat.github.io/NeIn/ |
|
|
- **Point of Contact:** [Nhat-Tan Bui](mailto:tanb@uark.edu) |
|
|
|
|
|
### Dataset Summary |
|
|
NeIn is the first large-scale for studying negation in text-guided image editing. It comprises 366,957 quintuplets, i.e., source image, original caption, selected object, negative sentence, and target image in total, including 342,775 queries for training and 24,182 queries for |
|
|
benchmarking image editing methods. |
|
|
|
|
|
### Dataset Structure |
|
|
- `COCO` (img): The source image from MS-COCO. |
|
|
- `T_original` (str): The original caption from COCO of that image. |
|
|
- `T_generated` (str): The addition instruction (e.g., "Add a couch.") for generate NeIn's sample and extract selected object. |
|
|
- `T_negative` (str): The negative instruction. |
|
|
- `NeIn` (img): The target image for negative instruction. |
|
|
|
|
|
In the context of image editing, given "NeIn" sample and "T_negative" clause, the ground truth corresponds to the “COCO” image. Please refer to our [paper](https://arxiv.org/abs/2409.06481) for more details. |
|
|
|
|
|
## Citation |
|
|
If you find this dataset useful, please consider citing our paper: |
|
|
``` |
|
|
@article{bui2024nein, |
|
|
author={Bui, Nhat-Tan and Hoang, Dinh-Hieu and Trinh, Quoc-Huy and Tran, Minh-Triet and Nguyen, Truong and Gauch, Susan}, |
|
|
title={NeIn: Telling What You Don't Want}, |
|
|
journal={arXiv preprint arXiv:2409.06481}, |
|
|
year={2024} |
|
|
} |
|
|
``` |
