This is the official experimental result of the undergraduate thesis of Qi as a supplementary material, please notice that this webpage is progressively in developing.
Originally, this thesis included some results and summary of the papers fellowing.
And their pipeline of implementation.
> Diffusion Model as good fine-tuning: DreamBooth, ControlNet, (and LoRA).
> Diffusion Model toward highly editive model: Prompt-to-prompt, Null-Text Optimization, Instruct Pix2Pix
> NeRF as representation for quick and high quality: NeuS, Instant-NGP.
> NeRF toward cross-modality of generation: DreamFusion, DreamBooth3D, RealFusion, Instruct-NeRF2NeRF.
Methodology of a dataset
USTC-2410 Dataset is a dataset which mainly recording of 3d things from 2-410 room of the middle campus of USTC as well as the beautiful USTC campus.
Feel free to pull request or put comments on this website for any good / constructive suggestions.
Always happy to listen to you and happy to work in this.
