Personalization as Inverse Planning: Learning Latent Design Intents for Agentic Slide Generation via Structural Denoising
For users of AI-based slide design tools, this work addresses the bottleneck of capturing fine-grained, page-level design intents without requiring verbose instructions or predefined templates.
This work tackles Page-level Slide Personalization (PSP) by formulating it as an inverse planning problem and proposing SPIRE, a framework that uses structural denoising and multi-agent reinforcement learning to infer latent design intents without tool-specific knowledge. Experiments show SPIRE outperforms existing methods, achieving higher personalization quality.
Slide design requires personalizing both deck themes and page layouts. Yet, current AI agent-based methods struggle with fine-grained, page-level design. Solely relying on prespecified templates or user verbose instructions, they fail to capture latent design intents, leaving Page-level Slide Personalization (PSP) unresolved. To close this gap, this work formulates PSP as an inverse planning problem. We propose to learn a design intent without assuming any knowledge of the specific executing tools (e.g., PowerPoint, Beamer) being used. However, relinquishing control over these tools makes the problem intractable to optimize end-to-end. To overcome this, we propose SPIRE, a principled framework to solve PSP approximately. By intentionally corrupting the visual structures of clean slides, SPIRE creates a verifiable task to denoise the corruption, whereby two agents learn to collaboratively refine executable designs via reinforcement learning (RL). We present a proof that structural denoising is a consistent surrogate for PSP, and that the multi-agent formulation strictly reduces policy gradient variance in RL. Extensive experiments demonstrate the superiority of SPIRE.