Skip to content

MiDaS 3.1

Pre-release
Pre-release
Compare
Choose a tag to compare
@thias15 thias15 released this 24 Dec 02:18
· 2 commits to master since this release
v3_1
1645b7e
  • New models based on 5 different types of transformers (BEiT, Swin2, Swin, Next-ViT, LeViT)
  • Training datasets extended from 10 to 12, including also KITTI and NYU Depth V2 using BTS split
  • Best model, BEiTLarge 512, with resolution 512x512, is on average about 28% more accurate than MiDaS v3.0
  • Integrated live depth estimation from camera feed
Assets 13