search

boxuLibrary / drt-vio-init

The official repository of our CVPR2023 paper "A Rotation-Translation-Decoupled Solution for Robust and Efficient Visual-Inertial Initialization".
GNU General Public License v3.0
155 stars 18 forks source link

Evaluation of gyr bias #21

Open ZoYooJy opened 2 weeks ago

ZoYooJy commented 2 weeks ago

According to the provided code, the groundtruth of the gyroscope bias comes from the average value in the dataset EuRoC file "state_groundtruth_estimate0/data/". However, reference[The EuRoC micro aerial vehicle datasets] points out that the gyroscope bias values provided in the EuRoC datasets are estimated using a classical maximum likelihood state estimator: "To determine these parameters, we used a classical maximum likelihood state estimator." Can you provide a detailed description of the relevant issues? Thanks a lot.

HeYijia commented 2 weeks ago

Is your question "How is the ground true value of gyro bias in the dataset obtained"? You can refer to the Kalibr. They estimated the bias by aligning the motion capture trajectory and the gyroscope measurement. paper link

ZoYooJy commented 2 weeks ago

Thank you for your reply!

---- Replied Message ---- | From | @.> | | Date | 07/01/2024 14:30 | | To | boxuLibrary/drt-vio-init @.> | | Cc | ZoYooJy @.>, Author @.> | | Subject | Re: [boxuLibrary/drt-vio-init] Evaluation of gyr bias (Issue #21) |

Is your question "How is the ground true value of gyro bias in the dataset obtained"? You can refer to the Kalibr. They estimated the bias by aligning the motion capture trajectory and the gyroscope measurement. paper link

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>