Benjamin Ummenhofer

The 128 is just for scaling the network output to roughly match the scale of the training data. The network can also learn this or the network initialization could be...

The radius is an important parameter for the collision with the boundary. Another thing that can help is to increase the number of frames for which losses are computed. For...

Sorry for the late reply. Unfortunately, there are no plans for float64 support for that op in Open3D. Is float64 needed for generating the data?

The normal is defined by the triangle the particle has been sampled from. https://github.com/isl-org/DeepLagrangianFluids/blob/d651c6fdf2aca3fac9abe3693b20981b191b4769/datasets/create_physics_scenes.py#L141

The radius should be chosen such that there is a reasonable number of neighbors on average (30-40 for our data). If the timestep is smaller, I guess the number of...

No, we did not do experiments with just two particles. The network handles collision well and in most of our scenes there are no problems but it can happen that...

To check if the network can produce larger values you can try to overfit to simulation frames with larger correction values. If that works then maybe increasing the importance of...

Just take 2 or 3 frames and see if the network can overfit to the large values in these frames. To make large corrections more important you can try changing...

position change, velocity, and acceleration are all related. You can have a look at this paper. They have experimented with different update schemes https://cgl.ethz.ch/disclaimer.php?dlurl=/Downloads/Publications/Papers/2015/Jeo15a/Jeo15a.pdf

I don't think that it is impossible for the model. Small timesteps usually make it easier for solvers. I would really just try to train on a dataset with just...