Amazing work on this project!
I noticed that the model runs out of memory when processing larger scenes with 30+ images. Are there any plans to support multi-GPU inference/training or other memory optimization strategies for large-scale scene reconstruction?
Amazing work on this project!
I noticed that the model runs out of memory when processing larger scenes with 30+ images. Are there any plans to support multi-GPU inference/training or other memory optimization strategies for large-scale scene reconstruction?