遇见数据集

sKT-Ai-Labs/W-IMG

收藏
Hugging Face2026-05-24 更新2026-05-31 收录
官方服务:

资源简介:

WIMG数据集是一个用于下一代图像生成的多模态图像和视频数据集合,由SKT AI LABS策划。该数据集包含100,000到1,000,000个原始图像,这些图像未经标注,专为WIMG模型设计,并面向开源社区开放。数据经过清理,确保无深度伪造内容,且不包含任何私人纪录片,涵盖了整个WIMG数据库。数据集语言为英语,适用于图像生成和计算机视觉任务。

The WIMG dataset is a multi-modal image and video data collection curated for next-generation image generation by SKT AI LABS. It contains between 100,000 and 1,000,000 raw images without captions, specifically designed for the WIMG model and open-sourced for the community. The data is cleaned, free from deepfakes, excludes any private documentary content, and encompasses the entire WIMG database. The dataset is in English and is suitable for image generation and computer vision tasks.

提供机构:
sKT-Ai-Labs
二维码
社区交流群
二维码
科研交流群
商业服务