xinshuoweng / Mono3DPLiDAR

Official PyTorch Implementation for "Monocular 3D Object Detection with Pseudo-LiDAR Point Cloud", ICCVW 2019
http://www.xinshuoweng.com/
131 stars 27 forks source link

question about dataloader #7

Open JianWu313 opened 3 years ago

JianWu313 commented 3 years ago

hello,is is possible to not to generate the huge pseuso-lidar point file(.bin). I have just write a function which transform the image to point cloud with the help of SOTA monocular-depth estimator. where should I put this function to avoid load the pretrained weight mutiple times when load the image-data by training.