Code release?

[petertheprocess]

Hello, impressive work, and congratulations on being accepted by TRO! I was wondering when the code will be open-sourced. Looking forward to it!

[ziweiWWANG]

Hello, impressive work, and congratulations on being accepted by TRO! I was wondering when the code will be open-sourced. Looking forward to it!

Hi Peter thank! The code and dataset will be released soon!

[maxrobic]

Hello! Thank you for this very clear paper and congratz for this achievement! Will the code be available on C++? Thanks a lot, I'm also looking forward to it :) Best

[ziweiWWANG]

Hello! Thank you for this very clear paper and congratz for this achievement! Will the code be available on C++? Thanks a lot, I'm also looking forward to it :) Best

Thank you! Yes it will be on C++.

[ziweiWWANG]

We are currently working on making the codebase cleaner and more accessible, which will take some time.

Please find the early release in https://drive.google.com/drive/folders/1Z3G4i4RDG_SjCmTIxGdwXQaOa_mpX-il?usp=sharing and follow the README file to run the code. Please feel free to ask if you have any other questions :)

[ccaccavella]

Hello, great work! Will the next release include data processing directly from live camera? Thanks.

[ziweiWWANG]

Hello, great work! Will the next release include data processing directly from live camera? Thanks.

Hi there, Thanks! This is not currently part of our plans, but we may consider extending it to a live demo in the future.

[ccaccavella]

Hi,

Thank you for you answer! Though you don't currently support live demo situation, do you expect the algorithm to perform comparably to what are the results on recorded data (at least for simple settings)? Thank you.

Best regards, Caterina Caccavella

On Thu, Oct 17, 2024, 01:24 Ziwei Wang @.***> wrote:

Hello, great work! Will the next release include data processing directly from live camera? Thanks.

Hi there, Thanks! This is not currently part of our plans, but we may consider extending it to a live demo in the future.

— Reply to this email directly, view it on GitHub https://github.com/ziweiWWANG/AEB-Tracker/issues/3#issuecomment-2418131198, or unsubscribe https://github.com/notifications/unsubscribe-auth/AT5UTBCU7VAO7KTQXMKDJ2DZ33YSTAVCNFSM6AAAAABOTZDFUCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIMJYGEZTCMJZHA . You are receiving this because you commented.Message ID: @.***>

[ziweiWWANG]

Hi, Thank you for you answer! Though you don't currently support live demo situation, do you expect the algorithm to perform comparably to what are the results on recorded data (at least for simple settings)? Thank you. Best regards, Caterina Caccavella

Hi Caterina,

Thank you. The short answer is yes if using appropriate parameters. To enhance live tracking, consider adding sliders or interactive controls that allow users to adjust parameters dynamically. This will provide real-time feedback and improve the overall experience.

[petertheprocess]

Hi ZiWei,

Thanks for sharing the code, I tried it but it runs very slow (less than 10 FPS) in my workstation. Is there some slow motion configuration set by default? Under what conditions and how do you measure speed? What do u think might be the issue? Thanks a lot.

https://github.com/user-attachments/assets/198470f2-9308-4e47-bf44-a601d6ecd65b

[ziweiWWANG]

Hi ZiWei,

Thanks for sharing the code, I tried it but it runs very slow (less than 10 FPS) in my workstation. Is there some slow motion configuration set by default? Under what conditions and how do you measure speed? What do u think might be the issue? Thanks a lot.

_org.wezfurlong.wezterm_20241024131920.mp4

Hi Peter,

You can make it faster by building in release mode (following README.md) It will generate the same video as the supplementary YouTube video. You can control the video framerate by changing the publish_framerate in the config file.

:)

    mkdir build
    cd build
    cmake -DCMAKE_BUILD_TYPE=Release ..
    cmake --build . -j8