Precise Imaging Guided Text-to-Image Diffusion Model
![截屏2024-05-20 09 56 20](https://private-user-images.githubusercontent.com/16849192/331921316-b58c558c-d6b9-400f-b1ef-1b589364ff68.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3MjU2MjUzMzYsIm5iZiI6MTcyNTYyNTAzNiwicGF0aCI6Ii8xNjg0OTE5Mi8zMzE5MjEzMTYtYjU4YzU1OGMtZDZiOS00MDBmLWIxZWYtMWI1ODkzNjRmZjY4LnBuZz9YLUFtei1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFWQ09EWUxTQTUzUFFLNFpBJTJGMjAyNDA5MDYlMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjQwOTA2VDEyMTcxNlomWC1BbXotRXhwaXJlcz0zMDAmWC1BbXotU2lnbmF0dXJlPTA0ODYyYzNlM2VhYjIyNjUxZjczOTBkNzFlOGNjYThjZDZjODJkMjcyMjI5OWRhNGM1ZmY5ZDUxOTRiZTM5YTAmWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0JmFjdG9yX2lkPTAma2V5X2lkPTAmcmVwb19pZD0wIn0.x9o3BjM9XjTl-1YtBunUL2jP-VWJoXXPbYoKB-QF0tw)
Most text-to-image models currently lack the capability for precise image generation control, such as camera view, light direction, camera characteristic. Our work introduce complex imageing condition to control image generation details.
If paper is accepted by NIPS, code would release.