NeuralRecon
NeuralRecon copied to clipboard
zju3dv
Code for "NeuralRecon: Real-Time Coherent 3D Reconstruction from Monocular Video", CVPR 2021 oral
Hello author, I would like to consult what specific requirements the NeuralRecon project has for the server? Can two T4's run?
When I run this command "python demo.py --cfg ./config/demo.yaml" Some trouble happened, like this:  What's wrong with it? My environment: Ubuntu 18.04 Pytorch+Cuda: 1.9.0+cu111 Python: 3.7
(process_with_single_worker pid=11508) read from disk (process_with_single_worker pid=33896) read from disk 0%| | 0/1 [00:00
Hi, thanks for your wonderful job and readily code! I downloaded the raw 7-Scenes data from its official website, but I am confused about how to process it so that...
Thank for excellent repo. I have a bunch of images, it contains 20 images (full of 360 degree). Now I want to reconstruct 3D 360 for generating indoor house. My...
pycuda
pycuda.driver.CompileError: nvcc compilation of /tmp/tmp3sh8k726/kernel.cu failed [command: /usr/local/cuda-11.0/bin/nvcc --cubin -arch sm_86 -I/home/dell/anaconda3/envs/neucon/lib/python3.7/site-packages/pycuda/cuda kernel.cu] [stderr: nvcc fatal : Value 'sm_86' is not defined for option 'gpu-architecture' ]
The SyncBatchNorm function will stop when one of my GPUs runs into "no valid points". Does anyone have any idea? How to solve the problem? I replaced the original BatchNorm...
Why inference on ScanNet test-set didn't test the "scene 806"? It always stops at scene 805. I have checked several times. I have the "scene 806" in the ScanNet test-set....
The code is as follows ``` import torch from torchvision import models from torch.utils.data import DataLoader # 有 GPU 就用 GPU，没有就用 CPU device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu') print('device',...
Hi team: Has anyone tried to convert the NeuralRecon model to onnx format? I remove the DataParallel before conversion (the DataParallel is reported to cause onnx exportion problem) but still...
