TeleStyle V2: Beyond Content-Preserving Style Transfer with Self-Distillation and Distribution-Matching-Distillation

2026-06-16

AI Summary

Existing methods often struggle to maintain content consistency when handling complex content–style combinations, such as swapping between artistic and photorealistic styles, due to ambiguity in reference ordering. This work proposes TeleStyle V2, which addresses this issue by constructing diverse content–style triplets through self-distillation and integrating distribution-matching distillation with prompt enhancement. The approach effectively resolves order ambiguity while preserving the base model’s general-purpose editing capabilities. Notably, TeleStyle V2 is the first method to uniformly support all four combinations—realistic-to-realistic (RnR), realistic-to-stylized (RnS), stylized-to-realistic (SnR), and stylized-to-stylized (SnS)—achieving style transfer quality comparable to Gemini-3-Pro-image-preview and general image editing performance on par with Qwen-Image-Edit-2509-DMD.

Abstract

Given a content reference and a style reference, content-preserving style transfer requires the model to generate stylized outputs with content and style consistency. We introduced TeleStyle V1 to tackle this problem. However, TeleStyle V1 is trained with photorealistic content reference and artistic style reference, which makes it incapable to cope with artistic content reference and realistic style reference in most cases. In this paper, we designed a Self-Distillation data synthesis strategy to construct such triplets from TeleStyle V1. Trained with such self-distilled triplets, our TeleStyle V2 supports Content-Style references in the forms of Realistic-and-Realistic (RnR), Realistic-and-Stylized (RnS), Stylized-and-Realistic (SnR), Stylized-and-Stylized (SnS). In addition, we found Distribution Matching Distillation could preserve the general text-guided image editing capability of the foundation model and fix the content consistency degradation caused by SFT process. Through quantitative evaluations, our TeleStyleV2-QIE-2509-DMD performs at least on par with Qwen-Image-Edit-2509-DMD, demonstrating strong general image editing skills beyond content-preserving style transfer. We observed the content/style reference order confusion problem in TeleStyle V1 and further introduced prompt enhancer to solve it. TeleStyle V2 uses Qwen-Image-Edit's VLM encoder, Qwen2.5-VL-7B, to generate content prompt and style prompt for free. TeleStyle V2 could achieve comparable style transfer performance with state-of-the-art commercial model, gemini-3-pro-image-preview.

Problem

• style transfer
• content-preserving
• artistic content
• realistic style
• reference confusion

Innovation

• Self-Distillation
• Distribution Matching Distillation
• Content-Preserving Style Transfer
• Prompt Enhancer
• Vision-Language Model