ShunChengWu comments

Results 18 comments of


                                            ShunChengWu

[Bug] AttributeError: module 'trimesh.voxel' has no attribute 'creation'

same here. able to solve this by import creation myself ``` from trimesh.voxel import creation ```

Conv operator

I believe that this is a 2D Conv. The kernel number define both kernel_h, kernel_w at the same time. The convolutional operation in default has dimensions of {batch, channel, height,...

Error in setup.py

Pytorch tries to use c++ compiler to compile a c project. an easy hack for this is to set your CXX to gcc `CC=gcc CXX=gcc python setup.py build_ext --inplace` of...

make issue: g++: error: roi_pooling_op.cu.o: No such file or directory

I think the easiest way to solve `make.sh: line 13: nvcc: command not found` is to set the path of your CUDA within the lib/make.sh file @riadhayachi this should also...

Wrong camera poses in ScanNet

If I remember correctly, I simply removed the NaN poses. Apart from that I didn't do anything else.

Using SceneGraphFusion with custom dataset

Hi Mihir, the backend runs with a given set of nodes and their neighbor graph for the semantic scene graph estimation. You can ignore the front-end part, which is the...

Using SceneGraphFusion with custom dataset

Hi Mihir, The original depth from the 3RScan dataset is terrible. That's why I had to use rendered depth from the mesh. The way you disabled the depth is okay....

Question about trace mode when I run main.py

I am not sure... can you format your text better? I guess it's related to Pytorch.

When to release the code ?

If this can be used to generate scene graph with a complete indoor model rather than rgb-d sequence?

Please check out my 3DSSG repo (https://github.com/ShunChengWu/3DSSG) You will need to change the input scene to the same format as provided by 3RScan dataset or the segment map ply files...