AnyPhoto: Multi-Person Identity Preserving Image Generation with ID Adaptive Modulation on Location Canvas

Longhui Yuan

cs.CVMar 16, 2026

AnyPhoto: Multi-Person Identity Preserving Image Generation with ID Adaptive Modulation on Location Canvas

Authors: Longhui Yuan

Abstract

Multi-person identity-preserving generation requires binding multiple reference faces to specified locations under a text prompt. Strong identity/layout conditions often trigger copy-paste shortcuts and weaken prompt-driven controllability. We present AnyPhoto, a diffusion-transformer finetuning framework with (i) a RoPE-aligned location canvas plus location-aligned token pruning for spatial grounding, (ii) AdaLN-style identity-adaptive modulation from face-recognition embeddings for persistent identity injection, and (iii) identity-isolated attention to prevent cross-identity interference. Training combines conditional flow matching with an embedding-space face similarity loss, together with reference-face replacement and location-canvas degradations to discourage shortcuts. On MultiID-Bench, AnyPhoto improves identity similarity while reducing copy-paste tendency, with gains increasing as the number of identities grows. AnyPhoto also supports prompt-driven stylization with accurate placement, showing great potential application value.

Figures & tables

Figure 1: AnyPhoto is capable of generating identity-preserving, location-controlled, high-quality images conditioned on reference faces and a text prompt.

Figure 2: AnyPhoto. Location-Aligned Token Pruning constructs the input from the text prompt, noised image, and location reference. Reference face embeddings modulate the aligned tokens. Training uses conditional flow matching and face similarity losses.

Figure 3: Detailed components illustration of AnyPhoto.

1-people subset
Method	Identity Metrics			Generation Quality
Method	Sim(GT) $\uparrow$	Sim(Ref) $\uparrow$	CP $\downarrow$	CLIP-I $\uparrow$	CLIP-T $\uparrow$	Aes $\uparrow$
DreamO	0.454	0.694	0.303	0.793	0.322	4.877
OmniGen	0.398	0.602	0.248	0.780	0.317	5.069
OmniGen2	0.365	0.475	0.142	0.787	0.331	4.991
FLUX.1 Kontext	0.324	0.408	0.099	0.755	0.327	5.319

Table 1: Comparison results on MultiID-Bench. , , and denote the 1st/2nd/3rd best results. For Copy-Paste/Generation-Quality ranking, we only include cases with

\mathrm{Sim(GT)}>0.40

\mathrm{Sim(Ref)}>0.50

. marks ignored cases in the ranking.

Figure 4: Visual comparisons of AnyPhoto with baselines conditioned on 1/2/3/4 persons.

Variants	Note	Identity Metrics			Generation Quality
Variants	Note	Sim(GT) $\uparrow$	Sim(Ref) $\uparrow$	CP $\downarrow$	CLIP-I $\uparrow$	CLIP-T $\uparrow$	Aes $\uparrow$
Base	(a)	0.160	0.161	0.002	0.620	0.311	5.178
$\quad+$ location $\&$ pruning	(b)	0.486	0.821	0.483	0.756	0.304	4.604
$\quad+$ modulation	(c)	0.494	0.836	0.518	0.747	0.297	4.957
$\quad+$ replacement $\&$ aug.	(d)	0.215	0.278	0.071	0.734	0.323	5.387
$\quad+$ $\mathcal{L}_{fs}$ (AnyPhoto)	-	0.424	0.587	0.186	0.763	0.326	4.960

Table 2: Ablation study of AnyPhoto on MultiID-Bench 3-and-4-person subset.

Panel of Figure 5 — Figure 5: Qualitative Ablations. The first row shows generations without a specified style, while the second row demonstrates outputs conditioned on a “Line Art” style.

Panel of Figure 6 — Figure 6: AnyPhoto is capable of generating high-quality images across diverse styles while preserving identity and enforcing accurate location control.

Panel of Figure 7 — Figure 7: Failure cases example of extreme conditioning.

Appendix figures & tables4 assets

Supplementary material from the paper’s appendix.

Appendix

Variants	Identity Metrics			Generation Quality
Variants	Sim(GT) $\uparrow$	Sim(Ref) $\uparrow$	CP $\downarrow$	CLIP-I $\uparrow$	CLIP-T $\uparrow$	Aes $\uparrow$
All visible	0.353	0.414	0.113	0.746	0.302	4.764
Identity-Isolated (AnyPhoto)	0.424	0.587	0.186	0.763	0.326	4.960

Appendix

Table A : Effect of Identity-Isolated Attention.

Variants	Identity Metrics			Generation Quality
Variants	Sim(GT) $\uparrow$	Sim(Ref) $\uparrow$	CP $\downarrow$	CLIP-I $\uparrow$	CLIP-T $\uparrow$	Aes $\uparrow$
Fixed $p=0.5$	0.403	0.529	0.141	0.762	0.316	4.871
Curriculum (AnyPhoto)	0.424	0.587	0.186	0.763	0.326	4.960

Appendix

Table B : Effect of curriculum reference face replacement.

Panel of Figure 12 — Figure A : Sensitivity to the trade-off hyperparameter $\lambda$ and LoRA Rank $r$ .

Figure B : Additional qualitative results of AnyPhoto. Across diverse scenes and 1–4 identity conditions, AnyPhoto preserves each identity while enforcing accurate location control, producing coherent compositions with fewer copy-paste artifacts.

Identity-Preserving Image Generation

Identity-preserving image generation creates or edits images while retaining a person's recognizable features across changes in pose, style, or scene.

See all 48 papers

WithEveryone: Unified Planning and Identity Grounding for Group Image Generation
Aug 20, 2026Hengyuan Xu, Qixun Wang, Yiji Cheng +5Subject-Driven Image Generation Image Generation
1Fudan University · 2Hunyuan, Tencent · 3The University of Hong Kong
CanvasComposer: Personalized Group Photo Generation via a Multi-Reference Canvas
Oct 23, 2025Gordon Guocheng Qian, Ruihang Zhang, Tsai-Shien Chen +11Subject-Driven Image Generation Image Generation
Snap Inc., USA · University of Toronto, Canada · University of California, Merced, USA +1
Diff-ID: Identity Consistent Facial Image Generation and Morphing via Diffusion Models
Jul 27, 2026Taimoor Rizwan, Sara Atito, Muhammad Awais +2Image Generation Identity-Preserving Image Generation
Centre for Vision, Speech and Signal Processing (CVSSP) University of Surrey Guildford, United Kingdom · School of Artificial Intelligence and Computer Science Jiangnan University Wuxi, China

Aug 20, 2026J/K move · Enter open · S save

WithEveryone: Unified Planning and Identity Grounding for Group Image Generation

Hengyuan Xu, Qixun Wang, Yiji Cheng, Miles Yang+4

1Fudan University · 2Hunyuan, Tencent · 3The University of Hong Kong

Subject-Driven Image Generation Image Generation Compositional Image Generation Identity-Preserving Image Generation Controllable Image Generation

Open paper AskPDF arXiv

Identity-preserving image generation becomes increasingly unreliable when a scene must contain many specified people. Beyond retaining each identity, the model must bind every reference to a distinct person and location, while training-time identity losses must establish correspondence among several noisy predicted faces. We introduce WithEveryone, a unified framework for generating group images up to ten reference identities. WithEveryone injects each selected identity as an addressed token, predicts a structured identity--layout plan, and renders the plan as a visual condition. Its key objective, Layout-Grounded ID Loss, uses annotated face regions to supervise the intended identities directly, avoiding unstable embedding-based face matching; ID Representation Forcing additionally trains a prediction for each identity before image synthesis. On an identity-disjoint benchmark, WithEveryone achieves the highest target-context identity similarity, improving face similarity from 0.462 for GPT-Image-2 to 0.499, while reducing copy-paste artifacts from 0.169 to 0.055. It further covers 97.3% of the requested identities with a duplicate rate of only 2.8%. These results show that explicit identity--layout grounding enables identity-preserving generation to scale to larger groups without relying on direct reference-face copying.

Abstract

Identity-Preserving Image Generation

Explore similar work

WithEveryone: Unified Planning and Identity Grounding for Group Image Generation

CanvasComposer: Personalized Group Photo Generation via a Multi-Reference Canvas

Diff-ID: Identity Consistent Facial Image Generation and Morphing via Diffusion Models