Back to datasets
Dataset assetOpen Source CommunityImage SegmentationHuman Instance Segmentation
chuonghm/MaGGIe-HIM
The MaGGIe dataset is a training and benchmark collection for instance‑aware alpha portrait matting in images and videos, specifically targeting mask‑guided matting tasks. Developed during Adobe Research’s 2023 Summer Internship and accepted at CVPR 2024, it serves applications such as image segmentation, instance matting, portrait matting, video matting, guided matting, and human matting.
Source
hugging_face
Created
Nov 28, 2025
Updated
Jun 14, 2024
Signals
186 views
Availability
Linked source ready
Overview
Dataset description and usage context
MaGGIe: Mask Guided Gradual Human Instance Matting
Dataset Overview
- Name: MaGGIe – Human Instance Image and Video Matting
- Task Category: Image Segmentation
- Tags:
- matting
- instance matting
- image matting
- video matting
- guidance matting
- human matting
- License: CC‑BY‑NC‑4.0
Detailed Description
- Project: A training and benchmark dataset for binary‑mask‑guided gradual human instance matting in images and videos.
- Status: Accepted at CVPR 2024.
Authors
- Chuong Huynh
- Seoung Wug Oh
- Abhinav Shrivastava
- Joon‑Young Lee
Citation
@inproceedings{huynh2024maggie,
title={Maggie: Masked guided gradual human instance matting},
author={Huynh, Chuong and Oh, Seoung Wug and Shrivastava, Abhinav and Lee, Joon-Young},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={3870--3879},
year={2024}
}
Need downstream help?
Pair the dataset with AI analysis and content workflows.
Once the source passes your review, move straight into summarization, transformation, report drafting, or presentation generation with the JuheAI toolchain.