3D Object Detection in Adverse Conditions?

Question

3D Object Detection in Adverse Conditions?

MartinHahner opened this issue 3 years ago · comments

Hi @keenan-burnett, great dataset! 🎉

Are you planning to release 3D Object Detection labels also for sequences containing rain and snowfall?
In your paper you e.g. mention Boreas-Objects-V1, so will there be a planned V2 release?

Keenan Burnett · Answer 1 · Wed Mar 30 2022 01:49:33 GMT+0800 (China Standard Time)

Hi Martin!

We're looking into this internally. Since we already have the data, it wouldn't be too much extra work to label some frames from the rain or snow sequences. I can't promise that there will be a V2 with these additions, but I'll say it could happen over the next year.

Mei Qi Tang · Answer 2 · Fri Mar 24 2023 23:32:19 GMT+0800 (China Standard Time)

Hi @keenan-burnett, any recent updates on this? Thanks!

Keenan Burnett · Answer 3 · Sat Mar 25 2023 02:20:31 GMT+0800 (China Standard Time)

Hi @meiiqi, we have plenty of interesting data taken during adverse weather but I currently don't have enough budget allocated for another Scale data annotation run. I know there are some researchers at my institute who are interested in pursuing this towards their winter autonomous driving research but it won't be happening any time soon.