Datasets:
File size: 2,943 Bytes
9b06dcc e865749 9b06dcc 0b06249 9b06dcc 0b06249 dd3f65e 0b06249 9b06dcc b6b719f e865749 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 | ---
annotations_creators:
- found
language:
- en
language_creators:
- found
license: apache-2.0
pretty_name: ObjectRemovalAlpha
size_categories:
- n<1K
source_datasets:
- lrzjason/ObjectRemovalAlpha
tags:
- object-removal
- image-inpainting
- image-editing
- text-to-image
task_categories:
- image-to-image
task_ids: []
dataset_info:
features:
- name: text
dtype: string
- name: ground_truth_image
dtype: image
- name: factual_image
dtype: image
- name: mask_image
dtype: image
splits:
- name: train
num_bytes: 57432291.0
num_examples: 20
download_size: 56530697
dataset_size: 57432291.0
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---
# Dataset Card for ObjectRemovalAlpha
## Dataset Description
- **Homepage:** https://huggingface.co/datasets/lrzjason/ObjectRemovalAlpha
- **Repository:** https://github.com/creative-graphic-design/huggingface-datasets/tree/main/datasets/ObjectRemovalAlpha
- **Original Dataset:** https://huggingface.co/datasets/lrzjason/ObjectRemovalAlpha
- **Hugging Face Dataset:** https://huggingface.co/datasets/creative-graphic-design/ObjectRemovalAlpha
- **Paper:** Not available in the original dataset release.
### Dataset Summary
ObjectRemovalAlpha is a small object-removal training dataset reformatted from `lrzjason/ObjectRemovalAlpha`. Each example contains a shared text prompt, a ground-truth image, a factual/source image, and a mask image.
The loader groups the original `_G`, `_F`, and `_M` files by numeric ID and exposes them as one row per object-removal example.
### Supported Tasks and Leaderboards
This dataset can be used for image editing, object removal, image inpainting, and mask-conditioned image generation experiments. No public leaderboard is bundled with the source dataset.
### Languages
The text prompts are in English.
## Dataset Structure
### Data Instances
```json
{
"text": "object-removal prompt",
"ground_truth_image": "<image>",
"factual_image": "<image>",
"mask_image": "<image>"
}
```
### Data Fields
- `text`: Prompt text shared by the original `_G`, `_F`, and `_M` files.
- `ground_truth_image`: Target image after removal.
- `factual_image`: Source image before removal.
- `mask_image`: Mask image indicating the removal region.
### Data Splits
| Split | Examples |
| --- | ---: |
| train | 20 |
## Dataset Creation
### Source Data
The data is loaded from the upstream Hugging Face dataset `lrzjason/ObjectRemovalAlpha` using `snapshot_download`.
### Licensing Information
The upstream dataset is marked Apache-2.0 on Hugging Face Hub.
### Citation Information
No dataset-specific paper or BibTeX citation was found in the upstream dataset release.
### Contributions
This Hugging Face dataset implementation was created by the creative-graphic-design organization to expose `lrzjason/ObjectRemovalAlpha` through a typed dataset loader.
|