confs

[chaojixiaosong]

No description provided.

[chaojixiaosong]

Hi, hello. May I ask if I have loaded a dataset of transparent water bottles that I took myself on your model? However, in the output‘ If there is nothing in the 'confs', may I ask if spann3r cannot reconstruct transparent or white objects, including objects in water

[HengyiWang]

Can you show some original images? - Generally speaking, I do not think the training of spann3r includes transparent objects, so it may not be able to handle that, but I found it does work in some underwater scenes.

[chaojixiaosong]

您能展示一些原始图像吗？- 一般来说，我不认为 spann3r 的训练包括透明对象，所以它可能无法处理，但我发现它在某些水下场景中确实有效。

Similar to these dozen or so images of transparent water bottles, the confidence generated in the output is shown in the confidence map above. Also, may I ask what kind of underwater scene you are referring to when certain underwater scenes are effective, and what are the requirements for the image? Or can you share your imaging results with me