The link in the dataset document failed

[zhanganny]

Hello，I‘m try to do work which needs the speech2gesture dataset. But it seems that the link provided in dataset.md is invalid now. Could you please provide another link? Thanks!

[shinydinosaur]

Hello! Unfortunately, I had to take it down since the university put size limits on our data online, but if you tell me exactly what you need (Video links? Annotations? Trained models?) I can upload them for you. I can also point you to the 3D-fied version of our dataset (https://talkshow.is.tue.mpg.de/ - if you use it, please cite us both) in case you are working with a 3D SMPL model (which I would recommend anyway if you are doing motion prediction). lmk and also feel free to send me a direct email to my Berkeley email address.

Cheers, Shiry

[zhanganny]

Hello! Unfortunately, I had to take it down since the university put size limits on our data online, but if you tell me exactly what you need (Video links? Annotations? Trained models?) I can upload them for you. I can also point you to the 3D-fied version of our dataset (https://talkshow.is.tue.mpg.de/ - if you use it, please cite us both) in case you are working with a 3D SMPL model (which I would recommend anyway if you are doing motion prediction). lmk and also feel free to send me a direct email to my Berkeley email address.

Cheers, Shiry

@shinydinosaur Thanks for you reply! I need all the link listed in the section "Download specific speaker data" and "Crop intervals containing frontal speaker" in speech2gesture/data /dataset.md. Include [the file frames_df_10_19_19.csv]、[a single or multiple speakers keypoints & frames tar file]、[the file containing all video links video_links.csv] and [intervals_df.csv]. Thanks again!

[sky24h]

@shinydinosaur Hi, thank you for your great work! I'm trying to reproduce this repo, and facing the same problem as this issue. Could you please also send me the listed files?

Include [the file frames_df_10_19_19.csv]、[a single or multiple speakers keypoints & frames tar file]、[the file containing all video links video_links.csv] and [intervals_df.csv].

Thank you in advance.

[JSHZT]

Hello! Unfortunately, I had to take it down since the university put size limits on our data online, but if you tell me exactly what you need (Video links? Annotations? Trained models?) I can upload them for you. I can also point you to the 3D-fied version of our dataset (https://talkshow.is.tue.mpg.de/ - if you use it, please cite us both) in case you are working with a 3D SMPL model (which I would recommend anyway if you are doing motion prediction). lmk and also feel free to send me a direct email to my Berkeley email address.

Cheers, Shiry

Hello, I'm trying to reproduce this working, please post the relevant files to me. Specifically, I actually need the motion-audio data pair. It would be better if I could send the pre-built data directly.You can also send via this email: [email protected]，thanks

[Polly9]

Hello, I am a Japanese graduate student majoring in information engineering. If you have an idea that you would like to implement using this dataset, could you please send us the complete dataset that can be downloaded "Download specific speaker data"? thanks.

[chence17]

Thanks for you reply! I need all the link listed in the section "Download specific speaker data" and "Crop intervals containing frontal speaker" in speech2gesture/data /dataset.md. Include [the file frames_df_10_19_19.csv]、[a single or multiple speakers keypoints & frames tar file]、[the file containing all video links video_links.csv] and [intervals_df.csv]. Thanks again!

@shinydinosaur Thanks for your work! I'm currently doing research about speech2gesture. I need all the files in the section "Download specific speaker data" and "Crop intervals containing frontal speaker" in speech2gesture/data /dataset.md. Include [the file frames_df_10_19_19.csv]、[a single or multiple speakers keypoints & frames tar file]、[the file containing all video links video_links.csv] and [intervals_df.csv]. Thanks again!

[ZDanielsResearch2]

Hello @shinydinosaur, I am also working on similar research, and access to this dataset would be very useful. Is it possible to obtain a copy of the CSVs and related data files as mentioned by others in this thread? Thank you!

[kyleo-flwls]

Hi @shinydinosaur,

I'm also working on some research related to this, and would like to be able to run the inference code for the models you provided. Would it be possible for you to send a link to the data mentioned above from the old google drive link to me as well? You can send it to kyle.olszewski AT flawlessai DOT com .

Thanks in advance!

--

[skywalker00001]

Hello! Unfortunately, I had to take it down since the university put size limits on our data online, but if you tell me exactly what you need (Video links? Annotations? Trained models?) I can upload them for you. I can also point you to the 3D-fied version of our dataset (https://talkshow.is.tue.mpg.de/ - if you use it, please cite us both) in case you are working with a 3D SMPL model (which I would recommend anyway if you are doing motion prediction). lmk and also feel free to send me a direct email to my Berkeley email address. Cheers, Shiry

Hello, I'm trying to reproduce this working, please post the relevant files to me. Specifically, I actually need the motion-audio data pair. It would be better if I could send the pre-built data directly.You can also send via this email: [email protected]，thanks

Hi, did you get the data from the authors? If so, could you please also consider sending me a copy?

[shinydinosaur]

I will upload this in the new year and update the website accordingly.

On Wed, Dec 18, 2024 at 12:19 AM Skywalk3r @.***> wrote:

Hello! Unfortunately, I had to take it down since the university put size limits on our data online, but if you tell me exactly what you need (Video links? Annotations? Trained models?) I can upload them for you. I can also point you to the 3D-fied version of our dataset ( https://talkshow.is.tue.mpg.de/ - if you use it, please cite us both) in case you are working with a 3D SMPL model (which I would recommend anyway if you are doing motion prediction). lmk and also feel free to send me a direct email to my Berkeley email address. Cheers, Shiry

Hello, I'm trying to reproduce this working, please post the relevant files to me. Specifically, I actually need the motion-audio data pair. It would be better if I could send the pre-built data directly.You can also send via this email: @.***，thanks

Hi, did you get the data from the authors? If so, could you please also consider sending me a copy?

— Reply to this email directly, view it on GitHub https://github.com/amirbar/speech2gesture/issues/23#issuecomment-2550651533, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAMSVAQUTMEPLX2RUE2OLW32GEVZBAVCNFSM6AAAAABE2X6RQ6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNJQGY2TCNJTGM . You are receiving this because you were mentioned.Message ID: @.***>

[MAXDoger]

I will upload this in the new year and update the website accordingly. …

hello ,thanks for your great work. I also need the following dataset files for study and research~~,the provided link is invalid~

[the file frames_df_10_19_19.csv]、[a single or multiple speakers keypoints & frames tar file]、[the file containing all video links video_links.csv] and [intervals_df.csv].

where can i get the dataset ?