Totoro97 / ProLiF

Hi, Could you provide the processing code for the Shiny dataset (similar to llff_dataset.py)? Because I can't get effective results in lab, crest and other scenes. Thanks.

Hi, thanks for your interest in this project! We did not use the data format provided in the Shiny dataset and only use the images. We use the code in LLFF to process the Shiny dataset. (https://github.com/Fyusion/LLFF/blob/master/imgs2poses.py)

Hi, thanks for your interest in this project! We did not use the data format provided in the Shiny dataset and only use the images. We use the code in LLFF to process the Shiny dataset. (https://github.com/Fyusion/LLFF/blob/master/imgs2poses.py)

Ah, what I want to ask is if you can provide a shiny dataset file, similar to:

ProLiF/modules/dataset/llff_dataset.py

Line 60 in a1f8900

class LLFFDataset:

I can only get good results on the cd and giants scenes, and it's very bad on the lab scene.

Hello, here is our processed Shiny dataset file: https://www.dropbox.com/s/u8fmns6tx7jwtnu/shiny_llff.zip?dl=0`. I ran COLMAP by myself to get the camera poses because I found it hard to understand the meaning of the data format provided in the original Shiny dataset. This cases now should be successfully trained because they now follow the same data convension of LLFF dataset.

The example config file to train is confs/prolif.yaml.

Thank you!

Shiny Dataset