BundleSDF
BundleSDF copied to clipboard
NVlabs
[CVPR 2023] BundleSDF: Neural 6-DoF Tracking and 3D Reconstruction of Unknown Objects
Hi, After getting passed the setuptool issue, a new error arises while building the docker, as seen below: ``` > [17/19] RUN conda activate py38 && cd /kaolin && FORCE_CUDA=1...
Can you provide a tutorial on creating a runtime environment using native anaconda?
I tried to set up the docker and run the run_custom.py on the provided milk data with use_gui 0. However, after some running, the code crashes at some point. Since...
Hi there, I wanted to know if there is any way to debug the cpp code, like the Bundletrack code? For the python part I can use the vscode debugger,...
Hi, the project page displays the reconstruction result of the drill. However, when I attempt to reconstruct the drill using the HO3D data (as shown below), the black part appears...
尊敬的wen 我进行自建数据纹理生成出现点云数据加载错误，具体问题如下： (BundleSDF) a@robot:~/project/BundleSDF$ python run_custom.py --mode run_video --video_dir /data/dataset/output --out_folder /data/dataset/bundlesdf_output --use_segmenter 1 --use_gui 1 --debug_level 2 Warp UserWarning: Python 3.9 or newer is recommended for running Warp, detected...
哥儿 @wing-kit @wenbowen123, Awesome work indeed! I have few questions, would highly appreciate your response - * Will BundleSDF be able to properly built mesh representations of the object at...
when I running ``` python run_custom.py --mode run_video --video_dir debug/2022-11-18-15-10-24_milk --out_folder debug/bundlesdf_2022-11-18-15-10-24_milk --use_segmenter 1 --use_gui 1 --debug_level 2 ``` The image of the first frame is printed,but then error occurs...
I downloaded the offical data for milk. According to the documentation, I ran run_custom.py in both mode run_video and mode global_refine, and obtained textured_mesh.obj and texture.png. However, the generated texture...
Let's suppose I have a realsense camera which is getting RGBD images continuously, can I pass these images to BundleSDF to make the reconstruction at near realtime, without running it...
