Visual AutoRegressive modeling (VAR)

Visual AutoRegressive modeling (VAR), a new generation paradigm that redefines the autoregressive learning on images as coarse-to-fine “next-scale prediction” or “next-resolution prediction”, diverging from the standard raster-scan “next-token prediction”.

Data and Resources

Cite this as

Keyu Tian, Yi Jiang, Zehuan Yuan, Bingyue Peng, Liwei Wang (2024). Dataset: Visual AutoRegressive modeling (VAR). https://doi.org/10.57702/n3ylo3mf

DOI retrieved: December 2, 2024

Additional Info

Field Value
Created December 2, 2024
Last update December 2, 2024
Author Keyu Tian
More Authors
Yi Jiang
Zehuan Yuan
Bingyue Peng
Liwei Wang
Homepage https://var.vision