VGDiffZero: Text-to-image Diffusion Models Can Be Zero-shot Visual Grounders

• Main Authors: Liu, Xuyang, Huang, Siteng, Kang, Yachen, Chen, Honggang, Wang, Donglin
• Format: Journal Article
• Language: English
• Published: 03.09.2023
• Subjects: Computer Science - Computer Vision and Pattern Recognition

Large-scale text-to-image diffusion models have shown impressive capabilities for generative tasks by leveraging strong vision-language alignment from pre-training. However, most vision-language discriminative tasks require extensive fine-tuning on carefully-labeled datasets to acquire such alignmen...