Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.19271 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908727813079040 |
|---|---|
| author | Song, Xinyang Wang, Libin Wang, Weining Li, Zhiwei Sun, Jianxin Zheng, Dandan Chen, Jingdong Li, Qi Sun, Zhenan |
| author_facet | Song, Xinyang Wang, Libin Wang, Weining Li, Zhiwei Sun, Jianxin Zheng, Dandan Chen, Jingdong Li, Qi Sun, Zhenan |
| contents | Recent image generation approaches often address subject, style, and structure-driven conditioning in isolation, leading to feature entanglement and limited task transferability. In this paper, we introduce 3SGen, a task-aware unified framework that performs all three conditioning modes within a single model. 3SGen employs an MLLM equipped with learnable semantic queries to align text-image semantics, complemented by a VAE branch that preserves fine-grained visual details. At its core, an Adaptive Task-specific Memory (ATM) module dynamically disentangles, stores, and retrieves condition-specific priors, such as identity for subjects, textures for styles, and spatial layouts for structures, via a lightweight gating mechanism along with several scalable memory items. This design mitigates inter-task interference and naturally scales to compositional inputs. In addition, we propose 3SGen-Bench, a unified image-driven generation benchmark with standardized metrics for evaluating cross-task fidelity and controllability. Extensive experiments on our proposed 3SGen-Bench and other public benchmarks demonstrate our superior performance across diverse image-driven generation tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_19271 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | 3SGen: Unified Subject, Style, and Structure-Driven Image Generation with Adaptive Task-specific Memory Song, Xinyang Wang, Libin Wang, Weining Li, Zhiwei Sun, Jianxin Zheng, Dandan Chen, Jingdong Li, Qi Sun, Zhenan Computer Vision and Pattern Recognition Recent image generation approaches often address subject, style, and structure-driven conditioning in isolation, leading to feature entanglement and limited task transferability. In this paper, we introduce 3SGen, a task-aware unified framework that performs all three conditioning modes within a single model. 3SGen employs an MLLM equipped with learnable semantic queries to align text-image semantics, complemented by a VAE branch that preserves fine-grained visual details. At its core, an Adaptive Task-specific Memory (ATM) module dynamically disentangles, stores, and retrieves condition-specific priors, such as identity for subjects, textures for styles, and spatial layouts for structures, via a lightweight gating mechanism along with several scalable memory items. This design mitigates inter-task interference and naturally scales to compositional inputs. In addition, we propose 3SGen-Bench, a unified image-driven generation benchmark with standardized metrics for evaluating cross-task fidelity and controllability. Extensive experiments on our proposed 3SGen-Bench and other public benchmarks demonstrate our superior performance across diverse image-driven generation tasks. |
| title | 3SGen: Unified Subject, Style, and Structure-Driven Image Generation with Adaptive Task-specific Memory |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.19271 |