VI-Depth
VI-Depth copied to clipboard
Code for Monocular Visual-Inertial Depth Estimation (ICRA 2023)
I'm running run.py on the VOID500 dataset. How should I post-process the output .pfm file so that the pixel values correspond to the true depth values? The current pixel values...
what a great work ! I want to know when to openSource the trainning code, Thx ~
Hi, thank you for your excellent work. Could you tell me how much FLOPS is required?
Thanks for this very interesting work, most depth predictor inputs always have inter-frame flicker, causing "ga_depth" to flicker too, how to remove the flicker between depth images? it's useful ---...
Hello @thias15 / @dwofk , thanks for this very interesting work. I have one perhaps naive question related to the license. As you release the pretrained models on the VOID...
Only IMU data and RGB are required in the flow chart. Is sparse metric depth map required in input?