遇见数据集

SpinML: Customized Synthetic Data Generation for Private Training of Specialized ML Models

收藏
NIAID Data Ecosystem2026-05-02 收录
数据链接:
官方服务:

资源简介:

This is a sample dataset required for running experiments described in the SpinML paper. The sample dataset is based of on Husky dogs that are classified based on the task that they are doing like sitting, sleeping, eating and playing. The dataset consists of both real dataset and synthetic dataset (generated using a diffusion model), that are useful to check the privacy utility trade off as mentioned in the SpinML paper.

创建时间:
2025-01-31
二维码
社区交流群
二维码
科研交流群
商业服务