antonyvan

Results 13 comments of antonyvan

The loss is displayed as NaN during the training process

Hello, Yitong, I got the similar issues. have you solved it ? Regards Chuanmao

dataset folder

I have the same question. do you solve it

dataset folder

Thanks a million for your prompt help, It is clear. A quick question. in crop_small.py line 80 why only crop one scene ? scene_dirs = scene_dirs[:1] "sample_matterport.py" produce some empty...

dataset folder

Really appreciate your prompt and clear answer. 👍

negative value for gaussian.rgb

Hi, Gaussian splatting viewer works fine with ply files produced by official gaussian splatting code. The training with Mip-nerf360 scenes (your provided data) is pretty the same as your results...

negative value for gaussian.rgb

I checked the opacity of official ply files. it doesn't apply sigmoid. ![Screenshot from 2024-04-25 16-28-49](https://github.com/joeyan/gaussian_splatting/assets/40614822/516a6ced-2f44-4f04-90da-f53ca35e5ef0)

negative value for gaussian.rgb

Here is the data https://drive.google.com/file/d/1k1FsiSPzVD6qR8UYwNEbYrcoY2dmJ8iR/view?usp=sharing Please download it. once it is done, please let me know I will stop the sharing. By the way I modified your dataloader a little...

negative value for gaussian.rgb

Actually the data I am sharing with you is my custom scene : ship_share. The image above is test result, which looks the training fails. the setting I am using...

negative value for gaussian.rgb

Thanks, I will try these.

How to obtain poses and depth maps from DROID-SLAM

check the visualization part. it generate point cloud from disparity (1/depth)