SlideGen: Collaborative Multimodal Agents for Scientific Slide Generation

2025-12-04

AI Summary

Existing approaches to automated academic slide generation from scientific papers primarily rely on text summarization, neglecting visual design and layout planning—key aspects of multimodal reasoning. This paper introduces the first modular, agent-based framework integrating visual planning, leveraging collaborative multimodal large language models to jointly perform long-document understanding, logical outline generation, cross-modal semantic alignment between text and figures, adaptive slide layout assignment, speaker note synthesis, and visual feedback-driven optimization. The system outputs editable PPTX files and supports iterative refinement, markedly enhancing design awareness and content fidelity. Evaluated across multiple benchmarks, the generated slides achieve expert-level performance in visual quality, content accuracy, and readability—establishing a new state-of-the-art for automated academic presentation generation.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Generating academic slides from scientific papers is a challenging multimodal reasoning task that requires both long context understanding and deliberate visual planning. Existing approaches largely reduce it to text only summarization, overlooking the visual component and design intensive nature of slide creation. In this paper we introduce SlideGen, an agentic, modular, and visual in the loop framework for scientific paper to slide generation. SlideGen orchestrates a group of vision language agents that reason collaboratively over the document structure and semantics, producing editable PPTX slides with logical flow and compelling visual presentation. By integrating coordinated outlining, mapping, arrangement, note synthesis, and iterative refinement, our system consistently delivers slides of expert level quality. Across diverse benchmarks and strong baselines, SlideGen outperforms existing methods in visual quality, content faithfulness, and readability, positioning it as the new state of the art in automated slide generation. Our work establishes a foundation for design aware multimodal slide generation, demonstrating how agentic collaboration can bridge understanding and presentation in complex multimodal reasoning tasks.

Problem

• Generates academic slides from scientific papers using multimodal reasoning.
• Addresses visual planning and design overlooked by text-only summarization methods.
• Produces editable PPTX slides with logical flow and expert-level quality.

Innovation

• Collaborative multimodal agents for slide generation
• Modular visual-in-loop framework for PPTX creation
• Iterative refinement ensuring expert-level slide quality