1 paper · 2 filters
Yu Yuan, Xijun Wang, Yichen Sheng +3
Image generation today can produce somewhat realistic images from text prompts. However, if one asks the generator to synthesize a specific camera setting such as creating differen…