GenMAC: Compositional Text-to-Video Generation with Multi-Agent Collaboration

TL;DR

GenMAC introduces a multi-agent, iterative framework for compositional text-to-video generation, effectively handling complex scenes by decomposing tasks and collaboratively refining outputs, leading to state-of-the-art results.

Contribution

This work presents a novel multi-agent, iterative approach with specialized agents and self-routing to improve compositional text-to-video generation.

Findings

Achieves state-of-the-art performance on compositional text-to-video benchmarks.

Effectively decomposes complex tasks into simpler, role-specific agent collaborations.

Demonstrates significant improvements over existing models in handling complex scene composition.

Abstract

Text-to-video generation models have shown significant progress in the recent years. However, they still struggle with generating complex dynamic scenes based on compositional text prompts, such as attribute binding for multiple objects, temporal dynamics associated with different objects, and interactions between objects. Our key motivation is that complex tasks can be decomposed into simpler ones, each handled by a role-specialized MLLM agent. Multiple agents can collaborate together to achieve collective intelligence for complex goals. We propose GenMAC, an iterative, multi-agent framework that enables compositional text-to-video generation. The collaborative workflow includes three stages: Design, Generation, and Redesign, with an iterative loop between the Generation and Redesign stages to progressively verify and refine the generated videos. The Redesign stage is the most challenging stage that aims to verify the generated videos, suggest corrections, and redesign the text prompts, frame-wise layouts, and guidance scales for the next iteration of generation. To avoid hallucination of a single MLLM agent, we decompose this stage to four sequentially-executed MLLM-based agents: verification agent, suggestion agent, correction agent, and output structuring agent. Furthermore, to tackle diverse scenarios of compositional text-to-video generation, we design a self-routing mechanism to adaptively select the proper correction agent from a collection of correction agents each specialized for one scenario. Extensive experiments demonstrate the effectiveness of GenMAC, achieving state-of-the art performance in compositional text-to-video generation.