Vision as Unified Multimodal Generation

Benchmark Model Rank Results
monocular-depth-estimation-on-diode-relativeSenseNova-Vision#1AbsRel: 0.206delta1: 0.764
monocular-depth-estimation-on-eth3d-relative-depth-alignedSenseNova-Vision#1AbsRel: 0.043delta1: 0.974
monocular-depth-estimation-on-kitti-relativeSenseNova-Vision#1AbsRel: 0.059delta1: 0.959
monocular-depth-estimation-on-scannet-relative-depth-alignedSenseNova-Vision#1AbsRel: 0.039delta1: 0.98
monocular-depth-estimation-on-nyuv2-relativeSenseNova-Vision#2delta1: 0.981AbsRel: 0.04
computer-use-agents-on-screenspot-v2SenseNova-Vision#44Accuracy: 85.9
object-detection-on-coco-val2017SenseNova-Vision#132box AP: 53.7