search

alexklwong / unsupervised-depth-completion-visual-inertial-odometry

Tensorflow and PyTorch implementation of Unsupervised Depth Completion from Visual Inertial Odometry (in RA-L January 2020 & ICRA 2020)
Other
192 stars 24 forks source link

evaluate visual odometry #2

Closed HuangGuoHui110 closed 1 year ago

HuangGuoHui110 commented 3 years ago

Hi, there are only depth evaluation in this project, how do you evaluate your pose?

alexklwong commented 3 years ago

Hello, thank you for your interest, we measured the absolute trajectory error, relative pose error and relative rotation error Their definitions were included in our supplementary materials.

As we progress with code cleanup/release, full implementation and examples will come as well. But in the short term (this week) I can include the implementations of the error metrics in src/eval_utils.py

alexklwong commented 3 years ago

Hello, you can use the pose evaluation metrics inside src/eval_utils.py, please see https://github.com/alexklwong/unsupervised-depth-completion-visual-inertial-odometry/commit/12d8beb280bbc5cea8512f2b9c74e21c4db930b0.