search

Owen-Liuyuxuan / visualDet3D

Official Repo for Ground-aware Monocular 3D Object Detection for Autonomous Driving / YOLOStereo3D: A Step Back to 2D for Efficient Stereo 3D Detection
https://owen-liuyuxuan.github.io/papers_reading_sharing.github.io/3dDetection/GroundAwareConvultion/
Apache License 2.0
361 stars 76 forks source link

About lack of dataset #80

Closed Chain-Mao closed 1 year ago

Chain-Mao commented 1 year ago

For yolostereo3d, I lack a binocular dataset of real scenes. I would like to ask if it is possible to enter only the image of monocular completed labeling during the training process, similar to the object detection dataset of YOLO? I want to reduce the requirements for the dataset, is it feasible?

Owen-Liuyuxuan commented 1 year ago
  1. YoloStereo3D is a stereo detection framework. Stereo images are needed in both training and testing.
  2. GAC and other monocular 3D detection methods in this repo support detection in 3D with only one camera. You could have a try with them.
Chain-Mao commented 1 year ago
  1. YoloStereo3D is a stereo detection framework. Stereo images are needed in both training and testing.
  2. GAC and other monocular 3D detection methods in this repo support detection in 3D with only one camera. You could have a try with them.

Thanks for your reply. I have learned a lot.