CVPR 2021 ChaLearn LAP Large Scale Signer Independent Isolated SLR Challenge (RGB TRACK) Forum

Go back to competition Back to thread list Post in this thread

> how to decode depth videos into realworld depths?

How can I decode the depth videos (actually encoded as RGB videos) back to get their original depths (in millimeters, such as provided by Kinect 2 sensor?.

Thank in advance

Posted by: eliasyauri @ Feb. 8, 2021, 10:49 p.m.

Dear participant,
Raw data was converted to be in the range of [0, 255] using a dynamic normalization factor. Unfortunately, these normalization factors have not been stored during the conversion. We are not planning to share raw data at this stage of the challenge. Thus, participants should work with the data in its current format. Note that Depth data is not allowed in the RGB track (the track you posted this message). In case you want to exploit Depth data, you need to submit your results to the RGB+D track (

Posted by: juliojj @ Feb. 9, 2021, 5:30 p.m.
Post in this thread