Sample Test Format v1#

Note

This is the documentation of our on-disk storage format for per-sample test datasets. If you are looking to export Noisebase-compatible data from your renderer or write a Noisebase-compatible data loader for your framework, you’re in the right place.

If you want to use a dataset, our corresponding data loader manual should be more helpful. If you are looking for datasets using the format, check out our datasets page.

Follows almost the same format as our training data, except that we store one frame per zip file for test sequences. In practice, each array is missing the F dimension but is otherwise the same.

Supported loaders#

  • Pytorch: noisebase.loaders.torch.TestSampleLoader_v1

Default format definition is in conf/noisebase/format/test_sample_v1.yaml.

A YAML file describing a dataset using the format needs to give the following source parameters and directories:

src:
  samples: 8
  rendering_height: 1080
  rendering_width: 1920
  files: sampleset_v1/test8/{sequence_name}/frame{index:04d}.zip
  sequences:
    - name: bistro1
      frames: 160
    - name: bistro2
      frames: 160
    # and so on

name: "Sample Set v1 - Test8 Dataset"

# Where denoised images should be saved
# relative to the provided `save_dir`
output: test8/{sequence_name}/frame{index:04d}.png

# Where reference images should be saved
# relative to the data directory
reference: sampleset_v1/ref/test8/{sequence_name}/frame{index:04d}.png

# Where computed metrics should be saved
# relative to the provided `save_dir`
metrics: test8/{sequence_name}.json

# Optionally skip the first few frames during metric computations
# to let the temporal buffer fill up
warmup: 16

Check conf/noisebase/sampleset_test8_v1.yaml for an example.

Radiance#

Key

Description

Dimensions

DType

color

RGBE encoded sample radiance

[4, H, W, S]

uint8

exposure

Minimum and maximum exposure per frame
for RGBE decoding

[2]

float32

reference

Clean radiance

[3, H, W]

float32

Sample geometry#

Key

Description

Dimensions

DType

position

Sample position in world-space

[3, H, W, S]

float32

motion

Change of world-space sample position
from last frame

[3, H, W, S]

float32

normal

Sample normal in world-space

[3, H, W, S]

float16

diffuse

Diffuse colour of the sample’s material

[3, H, W, S]

float16

Camera data#

Key

Description

Dimensions

DType

camera_position

Position of the camera in world-space

[3]

float32

camera_target

A point in world-space in the center
of the image (where the camera is looking)

[3]

float32

camera_up

Vector in world-space that points straight
upwards in screen-space (what’s upwards for the camera)

[3]

float32

view_proj_mat

Matrix mapping from world-space to screen-space

[4, 4]

float32

proj_mat

Matrix mapping from camera-space to screen-space

[4, 4]

float32

crop_offset

Offset of the image crop from (0,0) in pixel coordinates

[2]

int32