GigaVideo-1: Advancing Video Generation via Automatic Feedback with 4 GPU-Hours Fine-Tuning

TL;DR

GigaVideo-1 introduces an efficient, annotation-free fine-tuning framework for video diffusion models that leverages automatic feedback to enhance multiple video quality dimensions with minimal computational resources.

Contribution

It presents a novel automatic feedback-based fine-tuning method that improves pre-trained video diffusion models without human annotations or large datasets.

Findings

Achieves about 4% performance improvement across evaluation dimensions.

Requires only 4 GPU-hours for fine-tuning.

No manual annotations needed, demonstrating high efficiency.

Abstract

Recent progress in diffusion models has greatly enhanced video generation quality, yet these models still require fine-tuning to improve specific dimensions like instance preservation, motion rationality, composition, and physical plausibility. Existing fine-tuning approaches often rely on human annotations and large-scale computational resources, limiting their practicality. In this work, we propose GigaVideo-1, an efficient fine-tuning framework that advances video generation without additional human supervision. Rather than injecting large volumes of high-quality data from external sources, GigaVideo-1 unlocks the latent potential of pre-trained video diffusion models through automatic feedback. Specifically, we focus on two key aspects of the fine-tuning process: data and optimization. To improve fine-tuning data, we design a prompt-driven data engine that constructs diverse, weakness-oriented training samples. On the optimization side, we introduce a reward-guided training strategy, which adaptively weights samples using feedback from pre-trained vision-language models with a realism constraint. We evaluate GigaVideo-1 on the VBench-2.0 benchmark using Wan2.1 as the baseline across 17 evaluation dimensions. Experiments show that GigaVideo-1 consistently improves performance on almost all the dimensions with an average gain of about 4% using only 4 GPU-hours. Requiring no manual annotations and minimal real data, GigaVideo-1 demonstrates both effectiveness and efficiency. Code, model, and data will be publicly available.