Vision transformers for the dense prediction on 360° drone images in rescue environments

Dense prediction for each frame of the 1.4k 360° video (30fps) at the drone exercise at the Ahrtal, EA4, Rech. Look at the bird at 1:37!!
Credits: Rene Ranftl, Alexey Bochkovskiy and Vladlen Koltun: Vision Transformers for Dense Prediction, ArXiv preprint, 2021.