Datasets:
image imagewidth (px) 494 1.8k | label class label 10
classes |
|---|---|
01 | |
01 | |
110 | |
110 | |
22 | |
22 | |
33 | |
33 | |
44 | |
44 | |
55 | |
55 | |
66 | |
66 | |
77 | |
77 | |
88 | |
88 | |
99 | |
99 |
People Clothing Segmentation Dataset
Dataset comprises 14,358 high-quality photos of 7,179 people of diverse genders wearing bathing suits, each paired with detailed segmentation masks for precise body-parts segmentation. It designed for semantic and instance segmentation tasks, this large-scale collection offers manually annotated labels, enabling robust training for deep learning models in human body analysis.
By leveraging this dataset, researchers can train high-precision segmentation models to accurately delineate human bodies in swimwear under various real-world conditions. - Get the data
Metadata for the dataset
.png?generation=1751537200417063&alt=media)
It significantly improves segmentation accuracy for clothing-aware models, providing high-resolution imaging and corresponding annotations to advance computer vision tasks.
Frequently Asked Questions
What annotations are included besides the segmentation masks?
In addition to paired RGB images and segmentation masks, the dataset includes metadata that can support more detailed computer vision experiments. Available attributes include participant ID, gender, body frame category (normal, athletic, or overweight), tattoo presence, and whether the image was captured using a mirror.
Who can benefit from this human segmentation dataset?
The dataset is useful for machine learning researchers, computer vision engineers, academic institutions, and companies developing vision-based applications. Teams working on semantic segmentation, human parsing, pose estimation, virtual try-on systems, fitness analysis, or person understanding can use the paired images, masks, and metadata to train and evaluate robust deep learning models.
What image formats are used?
RGB photographs are provided in JPG format, while segmentation masks are stored as PNG files. This combination preserves high-quality image data alongside lossless mask annotations.
💵 Buy the Dataset: This is a limited preview of the data. To access the full dataset, please contact us at https://unidata.pro to discuss your requirements and pricing options.
With Photoshop-crafted masks, dataset supports applications like pose estimation, object detection, and human parsing.
🌐 UniData provides high-quality datasets, content moderation, data collection and annotation for your AI/ML projects
- Downloads last month
- 166