The model generated by custom prompt words is not good

threestudio-project / threestudio

A unified framework for 3D content generation.

Apache License 2.0

6.32k stars 480 forks source link

The model generated by custom prompt words is not good #306

Open Dragonkingpan opened 1 year ago

Dragonkingpan commented 1 year ago

I tested some of my own prompt words using dreamfusion if, and after 10000 steps of iteration, the results were much worse than those given in the official project. I don't know what the reason is, I will list the pictures below for everyone to take a look. Can you guess what my prompt word is? 微信图片_20230915184733 微信图片_20230915184737 微信图片_20230915184721 The official teddy bear is better. 微信图片_20230915184741

bennyguo commented 1 year ago

According to my experience, these are reasonable results using the dreamfusion-if pipeline. You may try the Magic3D or ProlificDreamer pipelines for more detailed results.

SEJUNHONG commented 1 year ago

Hello @bennyguo ~ I also experimented with dreamfusion-if using the prompts in the paper and the following blurry-feeling products were output. For what reason did you infer that a model that works well in 2D text-to-image is like that in dreamfusion-if?