Developing and Deploying Your Custom Action Recognition Application Without Any AI Expertise Using NVIDIA TAO and NVIDIA DeepStream

Originally published at:

Build an action recognition app with pretrained models, the TAO Toolkit, and DeepStream without large training data sets or deep AI expertise.

1 Like

I was trying to recreate the pre-trained model. I trained the net with the same 5 classes (ride bike, walk, fall floor, run and push) and the default hyperparameters. Then I use this model as a pre-trained model and train again with different classes. However, the accuracy and F1 I get with my recreated model don’t even come close to the ones I get with the downloaded pre-trained model. I thought this could be cause the hyperparameters used to train the model were different from the ones I used, but I couldn’t find any information on how the pre-trained model was trained.
I would highly appreciate if anyone could provide me with this information. The hyperparameters I am referring to are the ones in the train_rgb_3d_finetune.yaml file, such as the learning rate, epochs, momentum, weight_decay, dropout ratio, steps, batch size etc…

Thank you!

1 Like

Thanks for a fine guide on how to use TAO.

I am using my own data to train a model with 2 categories, but the pictures a full HD - and a 12 picture sequence from each video.

When I try to train the model I get an error :“ERROR: Unexpected bus error encountered in worker. This might be caused by insufficient shared memory (shm).”

It looks like the docker image needs extra memory. Can you please advice me on how to do that?


Hi, I am trying run deepstream-3d-action-recognition-app on jetson NX and it running out of memory. May this app can not run on jetson nx right?

Hi @samvdh - Can you share mode details about your setup, please?

  • Are you running on a container?
  • What version of DeepStream are you using?
  • What version of JetPack


Hi, my issue is temporary solved at this topic Deepstream 3d action recognition app leak memory on jetson nx - #19 by samvdh. Thank you for your support.

@samvdh – Great to hear! Let us know if you have any other questions…

1 Like