Darren Tsai

I made a docker image for this repo - feel free to use it. https://github.com/darrenjkt/CenterPoint

Lyft doesn't annotate vehicles on the opposite side of the road. See a quote from the [supplementary material](https://openaccess.thecvf.com/content/CVPR2021/supplemental/Yang_ST3D_Self-Training_for_CVPR_2021_supplemental.pdf) of the ST3D paper. "The Lyft dataset does not annotate objects on...

The waymo license is a strict non-commercial license so if you want to download any models trained on waymo data, please send me an email at [email protected] with your name,...

Running the `curl http://0.0.0.0:4242` gives the following: ```MLflowYou need to enable JavaScript to run this app.%```

Yes that's correct

Yes I can. I usually log my artifacts from a separate mflow client to the `mlflow server` though. For the client, I usually specify the mlflow_tracking_uri in python before I...

When I do that, I get this error `mlflow.exceptions.MlflowException: The configured tracking uri scheme: 'postgresql' is invalid for use with the proxy mlflow-artifact scheme. The allowed tracking schemes are: {'https',...

Ah thanks I got the following error in the tracking server logs related to permissions on my AWS user. ``` botocore.exceptions.ClientError: An error occurred (AccessDenied) when calling the DeleteObject operation:...

Solved it! I was exporting the AWS keys in the local environment but starting the mlflow server with a different set of AWS keys. Starting the mlflow server using the...

Not particularly for class agnostic nms but we reduced the number of classes to only predict super classes and it helps to make detections less noisy. I'm thinking a DETR...