Thomas Müller

Can you try again after updating to the latest commit? ```sh git pull origin master git submodule update --recursive ```

Hi, no, the real-time SLAM pipeline is not released yet, although you can hook your own via the Python bindings, which already permit a dynamically growing dataset like in the...

Hi there; thank you for the PR & thorough description! Just a few minor comments: - Posix paths were intentional for portability reasons. You'd want to be able to zip...

Agree on all accounts. I'm with you RE `--out` being somewhat redundant, but people have set up workflows that use colmap2nerf under the hood, and I don't want those to...

Hi, CUDA 10.2 and Visual Studio 2019 in C++14 mode don't seem to play well together. Could you try upgrading CUDA to 11.5 or higher? I've updated the "Requirements" section...

Any chance you pulled the latest changes and did not re-compile the codebase?

Hi, the reason some of the dimensionalities are commented out is that they increase the compile time and memory cost by quite a lot -- for a feature that most...

@KamiCalcium I'm afraid that's an issue on [NerfAcc](https://www.nerfacc.com/en/latest/)'s side, not tcnn. That said, if you would like to reproduce instant-ngp's results, I'd warmly encourage using the official implementation as none...

Hi, based on testing on my end, CUTLASS seems to not handle problem sizes that exceed the 32-bit integer limit. From your logs, the input matrix has size `8306688 *...

Hi Susan, thank you (and @yanglixiao1994) very much for this contribution as well as its thorough derivation and testing! I'd like to take the time to properly review it, but...