Skip to content

Dataset · DesmondYMTang2024

Generated images with person annotations for LanSE

DesmondYMTang2024/Language-Grounded_Sparse_Encoder_Training

Dataset of 100K–1M examples featuring AI-generated images and human annotations for interpreting and analyzing generative models.

What's inside

The repository contains neural network-generated images and human annotations prepared by researchers from NUS, Tsinghua, and Stanford. The data volume ranges from 100 thousand to 1 million English records. Materials are intended for image classification and image-to-text generation tasks.

What you can do with it

  • Training Language-Grounded Sparse Encoders
  • Analyzing and interpreting content created by generative models
  • Fine-tuning image classification and text description generation models
Tasks
Image classification, Image → text
Languages
English
Size
100K–1M rows
Downloads this month
33.6K
Likes
1

How to load it

from datasets import load_dataset

ds = load_dataset("DesmondYMTang2024/Language-Grounded_Sparse_Encoder_Training")