maskfusion
maskfusion copied to clipboard
Dynamic object reconstruction quality
Hi, I was able to run MaskFusion with the real input sequences from Co-Fusion, but the quality of the reconstruction obtained for the dynamic objects did not seem to be so good.
Results for teddy-handover.klg:
Results for sliding-clock.klg:
Results for place-items.klg:
Is this expected? Is there something that I should change in order to get better results?