Closed ernlavr closed 2 years ago
I didn't test the code on Windows. But I guess you need to make sure CUDA is working. As for the running time, it takes about 2s to render a 480 × 270 image using V100.
Would the 2s estimate be for a whole video or a single frame? I'm looking to experiment with this framework for a research project but I'm looking for something that could render a single viewpoint frame in real-time, atleast under 100ms
It's for a single frame. I am not focusing too much on the runtime. However, there are many great works on making NeRF more efficient. One example is Instant Neural Graphics Primitives with a Multiresolution Hash Encoding.
@gaochen315
Thanks for your response and for the link! I had missed that one during my paper-crawl!
Hi all,
Has anyone tried running the code on Windows? Does the framework use any Linux specific libraries? Is there any information regarding the average time necessary to run the forward-pass with the pretrained models?
Best Regards -E