Lin Li

Results 19 comments of Lin Li

I think they use prior camera knowledge to set the camera parameters. From my understaning, they use azimuth (0-360), elevation around 30, tilt 0, camera distance to the zero of...

@walsvid About the original data from shapenetcore.v1, are they all resized to be put in a unit cube, which are of size (-0.5,0.5)^3? And why would you choose ellipsoid with...

Thanks for your update! @walsvid But I found one problem. The initial ellipsoid face is not set correctly. ![Screenshot from 2020-02-05 19-26-03](https://user-images.githubusercontent.com/3291902/73824076-a409aa80-484d-11ea-8629-e96e99670ba7.png) ![Screenshot from 2020-02-05 19-26-20](https://user-images.githubusercontent.com/3291902/73824093-ab30b880-484d-11ea-960e-13e74a645a2f.png) Could you correct it?

@walsvid Did you update your ground truth data processing method? And I have found you use queue with 64 maximum-size for data fetch. Why did you choose this when your...

Sorry, that's a long-time question I don't remember my choice. But I think all your solutions should work. For the second choice, you could add data checking code to do...

@sunshineatnoon Hi, did u find a good voxorigin setting? Theoratically, as my understanding, voxel origin could be anywhere, cause this parameter is only for translation of transform. And Translation is...

@shurans So did you mean that perpareDataTest will just produce the volume ground truth?

@shurans So the `readDepth` function is just suitable for your synthetic depth image, right?