• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 100 subscribers
  • Views 387 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TDA4VM: pytorch-jacinto-ai-devkit : can't download dataset from training guide

Steven Lin3
Prodigy 20 points
Part Number: TDA4VM

In pytorch-jacinto-ai-devkit Multi task Network document, it provides Optical flow repository and Motion repository to download.

But I can't go to the link to download the data for training.

  • 0
    Prodigy 20 points

    Hi,

    It's also failed if I use git clone for the repository. See the result below.

    I also double checked my environment with "git clone git.ti.com/.../pytorch-jacinto-ai-devkit.git". It's OK.

    ------------------------------------------------------------------------------------------------------------------------------------------

    $ git clone bitbucket.itg.ti.com/.../browse

    Cloning into 'browse'...

    fatal: unable to access 'bitbucket.itg.ti.com/.../': Could not resolve host: bitbucket.itg.ti.com

    $ git clone bitbucket.itg.ti.com/.../

    Cloning into 'cityscapes_motion_dataset'...

    fatal: unable to access 'bitbucket.itg.ti.com/.../': Could not resolve host: bitbucket.itg.ti.com

    $ git clone bitbucket.itg.ti.com/.../leftimg8bit_flow_farneback_confidence

    Cloning into 'leftimg8bit_flow_farneback_confidence'...

    fatal: unable to access 'bitbucket.itg.ti.com/.../': Could not resolve host: bitbucket.itg.ti.com

    ------------------------------------------------------------------------------------------------------------------------------------------

    $ git clone git.ti.com/.../pytorch-jacinto-ai-devkit.git

    Cloning into 'pytorch-jacinto-ai-devkit'...

    remote: Enumerating objects: 983, done.

    remote: Counting objects: 100% (983/983), done.

    remote: Compressing objects: 100% (973/973), done.

    remote: Total 1297 (delta 668), reused 25 (delta 0), pack-reused 314

    Receiving objects: 100% (1297/1297), 891.43 KiB | 271.00 KiB/s, done.

    Resolving deltas: 100% (771/771), done.

    ------------------------------------------------------------------------------------------------------------------------------------------

  • 0 in reply to Steven Lin3
    TI__Genius 11436 points

    Hi,

    We have not publicly released the repository cityscapes_motion_dataset which contains the optical flow and motion annotations. Currently the link points to our internal repository used for our experiments.

    However the source code of pytorch-jacinto-ai-devkit is available and you can see how we are doing the MultiTask training.

    You can apply it to your dataset of interest. If you have a specific technical question about it, we will be happy to answer.

    Best regards,

    Manu