[2511.11483] ImAgent: A Unified Multimodal Agent Framework for Test-Time Scalable Image Generation

Computer Science > Computer Vision and Pattern Recognition arXiv:2511.11483 (cs) [Submitted on 14 Nov 2025 (v1), last revised 28 Mar 2026 (this version, v4)] Title:ImAgent: A Unified Multimodal Agent Framework for Test-Time Scalable Image Generation Authors:Kaishen Wang, Ruibo Chen, Tong Zheng, Heng Huang View a PDF of the paper titled ImAgent: A Unified Multimodal Agent Framework for Test-Time Scalable Image Generation, by Kaishen Wang and 3 other authors View PDF HTML (experimental) Abstract:Recent text-to-image (T2I) models have made remarkable progress in generating visually realistic and semantically coherent images. However, they still suffer from randomness and inconsistency with the given prompts, particularly when textual descriptions are vague or underspecified. Existing approaches, such as prompt rewriting, best-of-N sampling, and self-refinement, can mitigate these issues but usually require additional modules and operate independently, hindering test-time scaling efficiency and increasing computational overhead. In this paper, we introduce ImAgent, a training-free unified multimodal agent that integrates reasoning, generation, and self-evaluation within a single framework for efficient test-time scaling. Guided by a policy controller, multiple generation actions dynamically interact and self-organize to enhance image fidelity and semantic alignment without relying on external models. Extensive experiments on image generation and editing tasks demonstrate that ...