search

OpenDriveLab / OpenLane

[ECCV 2022 Oral] OpenLane: Large-scale Realistic 3D Lane Dataset
Apache License 2.0
492 stars 46 forks source link

Why is there no positive detection in my new set of images? #65

Closed adilsonmedronha closed 1 year ago

adilsonmedronha commented 1 year ago

Thank you for your excellent work.

I'm currently working on predicting a set of images in which I do not have annotations.

These images are newly captured by a car camera. I'm only interested in inference

In the result folder there is no image with detected lanes, why model is not detecting lanes?

here some samples: image image

Command I ran: bash eval_demo.sh output: p: 0 fp: 0 fn: 0 no positive detection no ground truth positive finished process file precision: -1 recall: -1 Fmeasure: -1

ilnehc commented 1 year ago

@adilsonmedronha Thanks. There are many possible reasons for this. As the images seem relatively clear, I think the most likely cause could be incorrect extrinsics / intrinsics parameters. Could you check with this?