About this document
Multi-Modal Models for Crop Disease Images by ahmadreza.safara is a document available to read on EtoBox.
The document discusses the development of PhytoSynth, a multi-modal generative model for generating synthetic crop disease images, particularly focusing on watermelon diseases. It highlights the limitations of existing generative adversarial networks (GANs) and presents a novel benchmarking approach for computational requirements and prompt engineering. The research demonstrates that the Stable Diffusion model SD3.5M can efficiently generate high-quality synthetic images from a limited number of real sample
- Author
- ahmadreza.safara
- Language
- EN