cjwbw/vqfr

Blind Face Restoration with Vector-Quantized Dictionary and Parallel Decoder

Capabilities

Reference Images

Community model (estimated from hardware time)

Name	Type	Description	Default	Constraints
`image`*	string(uri)	Input image. Output restored faces and whole image.	`—`	—
`aligned`	boolean	Input are aligned faces.	`false`	—

imagerequiredstring

Input image. Output restored faces and whole image.

alignedboolean

Input are aligned faces.

Default: false

Version: ccd53a9a38ebUpdated: 7/25/2026140.6K runs