bbzh

Euclidean space to be converted to NDC space

1

Hi, According to Eq.(16)-(17) in the NDC [writeup](https://github.com/bmild/nerf/files/4451808/ndc_derivation.pdf) and implementation in [here](https://github.com/bmild/nerf/blob/20a91e764a28816ee2234fcadb73bd59a613a44c/run_nerf_helpers.py#L143), the Euclidean space to be converted to NDC space is chosen as the view frustum (or field of...

camera intrinsic seems wrong

3

Hi NuPlan team, I have been looking into the sensor data, and would like to ask two questions on the camera intrinsic: 1. Camera intrinsic matrix for all cameras are...

Training code release?

1

Hi, thanks for the great work! When will the training code be released?

weird image masks

2

Hi, Thanks for releasing the dataset. The images are often with weird masks; an example is shown below. Are these artifacts?

render only foregound in Blender

3

Hi Yifan, Thanks for the great code release. The paper shows the foreground-only rendering from Blender (see below), but the code only produces the final composed image. Would you please...

Inconsistency in VAE config

1

Hi, The shape_model parameters are different in the [VAE config](https://github.com/wyysf-98/CraftsMan/blob/2f9ff141f8c39b7d38fbeb3da050e2be99a33b20/configs/shape-autoencoder/l256-e64-ne8-nd16.yaml#L29) and the [diffusion config](https://github.com/wyysf-98/CraftsMan/blob/2f9ff141f8c39b7d38fbeb3da050e2be99a33b20/configs/image-to-shape-diffusion/clip-mvrgb-modln-l256-e64-ne8-nd16-nl6.yaml#L37). They differ in qkv_bias and use_downsample. Which version is used? In addition, the pretrained VAE checkpoint...