Infinite You (by Zsxkib)
Discover Infinite You (by Zsxkib), the AI model that transforms your portrait photos into any style while strictly preserving your facial identity.
🚀Function Overview
Preserves facial identity in portraits while generating new styles, backgrounds, and contexts based on text prompts.
Key Features
- Maintains subject's facial identity during transformations
- Two model versions: aes_stage2 (better aesthetics) and sim_stage1 (higher identity similarity)
- Optional realism and anti-blur enhancement LoRAs
- Pose control via secondary input image
- Advanced identity strength and guidance parameters
Use Cases
- •Creating themed profile pictures
- •Generating historical/fictional character portraits
- •Professional photo editing with consistent identity
- •Artistic style transfer for portraits
⚙️Input Parameters
id_image
stringUpload a portrait image containing a human face. For multiple faces, only the largest face will be detected.
prompt
stringDescribe how you want the generated image to look. Be specific about details, style, background, etc.
control_image
stringOptional: Upload a second image to control the pose/position of the face in the output
model_version
stringChoose the model version - 'aes_stage2' for better text-image alignment and aesthetics, 'sim_stage1' for higher identity similarity
width
integerOutput image width in pixels (recommended: 768, 864, or 960)
height
integerOutput image height in pixels (recommended: 960, 1152, or 1280)
num_steps
integerNumber of diffusion steps - higher values (30-50) give better quality but take longer
guidance_scale
numberHow closely to follow the prompt (higher = more prompt adherence, lower = more freedom)
seed
integerRandom seed for reproducible results (None generates a random seed)
enable_realism
booleanApply the realism enhancement LoRA for more realistic-looking results
enable_anti_blur
booleanApply the anti-blur LoRA to reduce blurriness in the results
output_format
stringChoose the format of the output image
output_quality
integerSet the quality of the output image for jpg and webp (1-100)
infusenet_conditioning_scale
numberAdvanced: Controls how strongly the identity image affects generation (lower values = less identity preservation)
infusenet_guidance_start
numberAdvanced: When to start applying identity guidance (0.0-0.1 recommended)
infusenet_guidance_end
numberAdvanced: When to stop applying identity guidance (usually keep at 1.0)
💡Usage Examples
Example 1
Input Parameters
{
"seed": 666,
"width": 864,
"height": 1152,
"prompt": "A sophisticated gentleman exuding confidence. He is dressed in a 1990s brown plaid jacket with a high collar, paired with a dark grey turtleneck. His trousers are tailored and charcoal in color, complemented by a sleek leather belt. The background showcases an elegant library with bookshelves, a marble fireplace, and warm lighting, creating a refined and cozy atmosphere. His relaxed posture and casual hand-in-pocket stance add to his composed and stylish demeanor",
"id_image": "https://bytedance-infiniteyou-flux.hf.space/gradio_api/file=/tmp/gradio/50d7c87565bc7cd462d6b2f50935b35de2329bcbae7754013e11f81c120b0042/yann-lecun_resize.jpg",
"num_steps": 30,
"model_version": "aes_stage2",
"enable_realism": false,
"guidance_scale": 3.5,
"enable_anti_blur": false,
"infusenet_guidance_end": 1,
"infusenet_guidance_start": 0,
"infusenet_conditioning_scale": 1
}Quick Actions
Technical Specifications
- Hardware Type
- L40S
- Run Count
- 7.5k
- Commercial Use
- Unknown/Restricted
- Platform
- Replicate