Open 05063112lcs opened 2 weeks ago
photometric_confidence = confidence_list[0]confidence_list[1] confidence_list[2]*confidence_list[3]
, you should change the confidence threshold as well.There are still some questions :1. The evaluation indicators are consistent with those described in your paper, can it mean that I have successfully reproduced, despite the noise in my point cloud data? 2. As for the threshold you just mentioned, I also saw it in your reply to others, but I still don't know which parameter should be modified, please kindly give me some guidance, thank you!
---Original--- From: "Tianqi @.> Date: Mon, Jul 1, 2024 19:13 PM To: @.>; Cc: @.**@.>; Subject: Re: [TQTQliu/ET-MVSNet] About DTU dataset (Issue #19)
When changed to photometric_confidence = confidence_list[0]confidence_list[1] confidence_list[2]*confidence_list[3], you should change the confidence threshold as well.
No noise removal was used for point cloud data.
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>
1.yes
--conf
in the script. When changed to photometric_confidence = confidence_list[0]confidence_list[1] confidence_list[2]*confidence_list[3], the confidence threshold --conf
should be smaller.I see a test dtu pbd.sh file in the code you distribute. This script file is not used in the entire reproduction process. Instead, I use the test dtu.sh file. I was testing on TNT data set today. Is the testing process of tnt adv very slow? My testing process has been almost 10h and has not been completed yet. On the contrary, tnt inter has been finished.
---Original--- From: "Tianqi @.> Date: Mon, Jul 1, 2024 19:31 PM To: @.>; Cc: @.**@.>; Subject: Re: [TQTQliu/ET-MVSNet] About DTU dataset (Issue #19)
1.yes
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>
Yes, it takes longer, but you can see some intermediate results.
Hello, I modified the conf indicator according to what you said, and the final acc, comp, and all indicators are very close to your paper, but the qualitative analysis results are still very different from those in your paper, no matter how small the settings are, there is still a big difference between them and your paper, I haven't changed your code, I don't know what the reason is, can you help solve it, the GPU I use is A5000, or the evaluation of the DTU dataset needs to run test
-pcd_dtu.sh instead of test_dtu.sh when I set the conf indicator to 0.1, the effect is as follows.
Hello, I have some questions when I reproduce your work, and I hope you can help answer them.