Clarification on the object detection pipeline #5052
Replies: 1 comment
-
@WilliamFlinchbaugh thank you for your interest in Autoware.
|
Beta Was this translation helpful? Give feedback.
-
@WilliamFlinchbaugh thank you for your interest in Autoware.
|
Beta Was this translation helpful? Give feedback.
-
Hi there, I'm pretty new so forgive me. I'm just super confused on how exactly the object detection pipeline works. Looking at the documentation, I see that it's multimodal, so it uses the camera and LiDAR data to create a 3d bounding box. However, I also see other packages like the YOLOX, ByteTrack, and image projection based fusion which don't seem to be used by default. Here are my main questions:
Are any of the packages used by the perception system based purely on camera data, or is it all either LiDAR or a combination of LiDAR and camera data?
What is the primary package responsible for detecting and classifying objects around the vehicle? For example, when autoware shows the bounding cylinder around a pedestrian, which package classifies the object as a pedestrian, and is the classifier based on only camera data?
If I wanted a 2d image output of the front camera with bounding boxes, what is the easiest way to go about it? I've seen the /perception/object_recognition/detection/rois0 topic seems to be what I want, but it doesn't seem to be active when I'm running autoware. Is this because the YOLOX package isn't active?
Any help or clarification is much appreciated, thank you!
Beta Was this translation helpful? Give feedback.
All reactions