0

SynthVLM: High-Efficiency and High-Quality Synthetic Data for Vision Language Models

SynthVLM uses advanced diffusion models to generate high-resolution images from captions, achieving state-of-the-art performance in vision question answering tasks with significantly reduced computational overhead and enhanced privacy.

Year
2024
Venue
arXiv 2024
Authors
10
Hosting
Abstract onlyARXIV-DEFAULT

Cite

Notes

Only stored in your browser.

Attribution

Abstract & full text
arxiv.org/abs/2407.20756v3ARXIV-DEFAULT
TL;DR
Semantic Scholar
Attribution policy →

Abstract

Recently, with the rise of web images, managing and understanding large-scale image datasets has become increasingly important. Vision Large Language Models (VLLMs) have recently emerged due to their robust vision-understanding capabilities. However, training these models requires vast amounts of data, posing challenges to efficiency, effectiveness, data quality, and privacy. In this paper, we introduce SynthVLM, a novel data synthesis pipeline for VLLMs. Unlike existing methods that generate captions from images, SynthVLM employs advanced diffusion models and high-quality captions to automatically generate and select high-resolution images from captions, creating precisely aligned image-text pairs. Leveraging these pairs, we achieve state-of-the-art (SoTA) performance on various vision question answering tasks, maintaining high alignment quality and preserving advanced language abilities. Moreover, SynthVLM surpasses traditional GPT-4 Vision-based caption generation methods in performance while significantly reducing computational overhead. Crucially, our method's reliance on purely generated data ensures the preservation of privacy, achieving SoTA performance with just 100k data points (only 18% of the official dataset size).

Authors

10