Zengyi Qin

Yes 11 recall positions are used to calculate the AP. The issue is also clarified in https://github.com/Zengyi-Qin/Weakly-Supervised-3D-Object-Detection/issues/8

In `evaluate_object.cpp` of your kitti evaluation code, was the const double N_SAMPLE_PTS in line 59 set to 41 or 11? It seems that the current cpp file that KITTI evaluation...

We would suggest you to train a customized base speaker tts model. Collect your own data and use any tts repo to train it. If you don't want to train...

Hey - Thanks for your suggestion. We will keep this in mind. Currently the whole team has a lot of TODOs, so we probably won't have time for this. But...

Hi - We are actually workinig on a new project that will opensource the data (with diverse emotion/styles) and code. Stay tuned!

Comparison will be added in Jan. We will update the website and paper

Chinese base speaker added. Please see demo_part1.ipynb

It's not in the video. It's in demo_part1.ipynb.

The env setup should be very basic and easy. What specific issue did you encounter?

Hi - I wonder if you have any follow-up questions. We will consider closing this issue if there's no follup-ups