Hi @shalinidemello
Thanks for your great work!
For the performance evaluation on BIWI dataset, there are two parts confusing me:
- where comes from the index file and why dont't we evaluate on these samples
- why there is a LinearRegression for each rotation angle used in testing to fit predictions and annotations, are we allowed to use annotations during testing?
I'm not very familiar with the task of HeadPose estimation, please forgive me if I misunderstood sth.
Hi @shalinidemello
Thanks for your great work!
For the performance evaluation on BIWI dataset, there are two parts confusing me:
I'm not very familiar with the task of HeadPose estimation, please forgive me if I misunderstood sth.