xygs1121

onnx inference problem

2

I convert the model to onnx using the model_to_onnx.py, with the args:--dataset nyuv2 --last_ckpt ./trained_models/nyuv2/r34_NBt1D.pth the model can be converted to model.onnx, the input shape is : NodeArg(name='rgb', type='tensor(float)', shape=[1,...

I think there are 2 bugs in the pull_item(self, index) function in coco.py when the annotation is empty, it will go to the else branch: if self.transform is not None:...

the mask in rendering result

7

I noticed in this function ```python res = render_func(view, gaussians, pipeline,...) ``` 它的结果res包含了"mask", 也注意到在diff-gaussian-rasterization的forward.cu中有渲染mask. 但是这个mask画出来并不像mask, 而是一些稀疏的点。 请问在evaluate中，计算和ground truth mask的IOU时能用res['mask']吗？ 还是需要用res['render']中>0的部分作为predicted mask? 期待回复 res["render"]  res['mask'] ![Screenshot from...

rendered image and gt image are different

2

when I overlap the gt image and the rendered image, I found they are not in the same location. I trained the truck for 10000 iterations, and use rendered.py to...

the mip360/garden dataset not work

1

我用的v2 branch, 因为cuda内存限制，garden images用的是downsample=8 在分割时click 了花，花盆，桌子，但是不管哪种情况分割出的都是地板。  但是从feature上看，它们的特征是不同的，请问这是哪里有问题呢？  相关代码： ```python fig = plt.figure() plt.title("please click the target") plt.axis("off") ax = fig.add_subplot(111) ax.imshow(view.original_image.permute(1,2,0).detach().cpu().numpy()) cid...