search

ZrrSkywalker / MonoDETR

[ICCV 2023] The first DETR model for monocular 3D object detection with depth-guided transformer
327 stars 31 forks source link

About depth map supervision #55

Open shanqiu24 opened 6 months ago

shanqiu24 commented 6 months ago

I want to know when supervising the depth map, does Object-wise Depth Labels mean that all pixels in the object 2D BBox are set to the same depth bin?