# bytedance/InfiniteYou

🔥 [ICCV 2025 Highlight] InfiniteYou: Flexible Photo Recrafting While Preserving Your Identity

Repository: https://github.com/bytedance/InfiniteYou
Canonical: https://ross.abutalabs.com/products/infiniteyou
Homepage: https://bytedance.github.io/InfiniteYou/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: face, flux, identity-preserving, image-editing, image-generation, personalization, text-to-image, diffusion, diffusers, pytorch, research, iccv2025, diffusion-transformer, dit
Last push: 2025-08-22T21:07:45+00:00

## Health v2 (maintenance only)
Score: 37/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 38, release rhythm 35, longevity 40
- inputs: {"age_days": 565, "days_push": 376, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2685, forks 286 (observed 2026-08-28T04:07:10.237626+00:00)

## What it is
InfiniteYou (InfU) is a research framework from ByteDance for identity-preserved text-to-image generation built on Diffusion Transformers like FLUX. It injects face identity features via InfuseNet with residual connections and multi-stage training to enable flexible photo recrafting while preserving a person's identity.

## Use cases
- generate new photos of a person from a reference face image
- recraft a portrait photo with a text prompt while keeping the face identity
- create identity-preserving avatars or headshots with text-to-image
- improve face similarity in FLUX-based image generation
- avoid face copy-pasting when generating images of a specific person
- plug identity features into an existing diffusion transformer pipeline
- run identity-preserved image generation in ComfyUI

## When to choose
- you need high-fidelity identity preservation with strong text-image alignment on FLUX/DiT models
- existing tools like IP-Adapter or PuLID-FLUX give insufficient identity similarity or poor aesthetics
- you want a plug-and-play identity injection component compatible with other methods
- you are doing research on identity-preserved generation with diffusion transformers

## When to avoid
- you need general-purpose image editing without identity constraints
- you have no GPU or cannot run large diffusion models locally
- you need a production-ready end-user app rather than a research codebase
- you work with non-face identity or object personalization only

## Facets
- artifact type: library
- maturity: active
- function: image-processing, machine-learning, deep-learning, stable-diffusion, llm-training
- domain: image-processing, deep-learning, computer-vision, artificial-intelligence
- platform: python
- tags: identity-preservation, text-to-image, diffusion-transformer, flux, face-generation, personalization, research-paper, diffusers, comfyui, gpu

## Member repositories
- bytedance/InfiniteYou (main) score 37

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:10.237626+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:17:49.607920+00:00, confidence not recorded.
  - readme: https://github.com/bytedance/InfiniteYou (fetched 2026-08-28T04:07:10.237626+00:00, sha ef4d438b47da)
  - homepage: https://bytedance.github.io/InfiniteYou/ (fetched 2026-08-29T09:59:55.787403+00:00, sha 2303b8df0272)
- Data as of 2026-08-30T08:39:29.467469+00:00.