no code implementations • 20 May 2024 • Wentao Wang, Xi Xiao, Mingjie Liu, Qing Tian, Xuanyao Huang, Qizhen Lan, Swalpa Kumar Roy, Tianyang Wang
MDT-AF incorporates an attention-based feature filtering mechanism into the patch embedding blocks and employs a coarse-to-fine process to mitigate the impact of low signal-to-noise ratio.
1 code implementation • 16 Dec 2023 • Wentao Wang, Xuanyao Huang, Tianyang Wang, Swalpa Kumar Roy
This paper explores the image synthesis capabilities of GPT-4, a leading multi-modal large language model.