Image-to-Image 3.0 - Portrait Photography
By referencing the facial features of a single real person in the input image and guiding with a text prompt, you can generate a variety of real-person portraits. This can be widely applied in interactive entertainment, portrait effects, and e-commerce marketing scenarios. Preset style keywords can be used as templates, or you can debug new template effects based on examples.
Reference: https://www.volcengine.com/docs/85128/1602212#wGHRz2FN
Price: 0.05 PTC/image
Request Parameters Header ParametersAuthorizationstringOptional Example Value:Bearer {{YOUR_API_KEY}} Query Parametersresponse_formatstringOptional Example Value:url Request Body application/jsonimage_inputstringRequired Width of the generated image
Default value: 1328
Value range: [512, 2048]
Height for generated image
Default value: 1328
Value range: [512, 2048]
High-definition processing effect. The higher the value, the clearer the face. It is recommended to use the default value.
Default value: 0.4
Value range: [0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1]
Facial beautification effect. The higher the value, the more noticeable the beautification effect. It is recommended to use the default value.
Default value: 0.3
Value range: [0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1]
Skin smoothing effect. The higher the value, the more noticeable the even skin tone and blemish removal effect. It is recommended to use the default value.
Default value: 0
Value range: [0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1]
Reference modes and value descriptions:
“creative”: Prompt mode (automatically switches to this mode when there is a prompt) (only references the character’s gender features)
“reference”: Full reference mode (automatically switches to this mode when there is no prompt) References more dimensions of the character (gender, clothing, hairstyle, etc.) and background
“reference_char”: Character reference mode, references more dimensions of the character (gender, clothing, hairstyle, etc.), does not reference the background
Manual input takes priority over automatic
Random seed, used as the basis for determining the initial state of diffusion. Default is -1 (random). If the random seed is the same positive integer and all other parameters are identical, the generated content will most likely be consistent.
Default value: -1
|
Image-to-Image 3.0 - Portrait Photography Task Acquisition
Price: Free
Request Parameters Header ParametersAuthorizationstringOptional Example Value:Bearer {{YOUR_API_KEY}} Query Parametersresponse_formatstringOptional Example Value:url Request Body application/jsonTask ID, the value of this field is returned by the submit task interface
String after JSON serialization
Currently supports watermark configuration and whether to return as an image link, which can be added in the return result.
LogoInfo
Watermark related information
return_url Optional bool Output whether to return an image link (the link is valid for 24 hours)
logo_info Optional LogoInfo Watermark information
add_logo Boolean No Whether to add a watermark. True to add, False not to add. Default is not to add
position Int No Position of the watermark, values are as follows:
0 - Bottom right
1 - Bottom left
2 - Top left
3 - Top right
Default is 0
language Int No Language of the watermark, values are as follows:
0 - Chinese (AI生成)
1 - English (Generated by AI)
Default is 0
opacity Float No Opacity of the watermark, value range 0-1, 1 means fully opaque, default is 0.3
|
