visual-genome
VisualGenome_VG_100K_1_and_2
Synthetic_Visual_Genome2
Synthetic Visual Genome 2 (SVG2)
A large-scale panoptic video scene graph dataset containing object labels, attributes, relationships, and instance-level segmentation masks.
Paper: Synthetic Visual Genome 2: Extracting Large-scale Spatio-Temporal Scene Graphs from Videos
Website: Synthetic Visual Genome 2
Versions
cleaned
The cleaned version has two sources: PVD (~593K videos) and SA-V (~43K videos).
SAM-3 outputs
We also provide the instance masks and… See the full description on the dataset page: https://huggingface.co/datasets/UWGZQ/Synthetic_Visual_Genome2.VisualGenome
Visual Genome Raw Data
This repository provides a convenient copy of the raw Visual Genome dataset files for easier access and downloading.
This is not an official Visual Genome release.The original dataset was created and released by the Visual Genome authors.Please refer to the official Visual Genome website and paper for authoritative information.
Dataset Summary
Visual Genome is a large-scale vision-language dataset with dense image annotations, including:
Images… See the full description on the dataset page: https://huggingface.co/datasets/jn12/VisualGenome.visual_genomeVisual Genome enable to model objects and relationships between objects.
They collect dense annotations of objects, attributes, and relationships within each image.
Specifically, the dataset contains over 108K images where each image has an average of 35 objects, 26 attributes, and 21 pairwise relationships between objects.visual_genomevisual_genomeVisual Genome (108,077 images). Canonical shards live under data/.