Excuse me, have anyone tested the data sets in euroc? I used to work in vins_ Mobile, but The average error of V1_01_easy running vins_mobile reached 0.305954m, and then in vins_mono , the average error is 0.044289. is there any wrong setting of vins_mobile? Because I only have three images to calculate the pose once, and then use all the IMU data to initialize successfully. I want to calculate the pose of two pictures once, but the initialization fails.
6s result. Evaluated with evaluate_ate.py provided by tum
Excuse me, have anyone tested the data sets in euroc? I used to work in vins_ Mobile, but The average error of V1_01_easy running vins_mobile reached 0.305954m, and then in vins_mono , the average error is 0.044289. is there any wrong setting of vins_mobile? Because I only have three images to calculate the pose once, and then use all the IMU data to initialize successfully. I want to calculate the pose of two pictures once, but the initialization fails. 6s result. Evaluated with evaluate_ate.py provided by tum
vins_mono result