yoxu515 / aot-benchmark

An efficient modular implementation of Associating Objects with Transformers for Video Object Segmentation in PyTorch
BSD 3-Clause "New" or "Revised" License
601 stars 108 forks source link

AOT/DeAOT perfs on MOSE dataset #33

Closed bhack closed 1 year ago

bhack commented 1 year ago

Just in the case you are interested: https://henghuiding.github.io/MOSE/

z-x-yang commented 1 year ago

I have noticed this work, a new challenging benchmark for VOS.

bhack commented 1 year ago

Yes but it is already a good baseline. A new challenge is also https://taodataset.org/workshop/cvpr23/#competition

z-x-yang commented 1 year ago

Yes, I'm watching on this challenge as well. Are you a organizer of this workshop?

bhack commented 1 year ago

Are you a organizer of this workshop?

No, but it seems interesting. The dataset is mainly: https://github.com/Ali2500/BURST-benchmark

bhack commented 1 year ago

If you are interested the MOSE competition is available now on CodaLab: https://codalab.lisn.upsaclay.fr/competitions/10703