Rgb noise models #357
Merged
Rgb noise models #357
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…itat-sim into redwood-noisy-depth
…itat-sim into redwood-noisy-depth
… redwood-noisy-depth
… redwood-noisy-depth
erikwijmans
reviewed
Nov 20, 2019
| @@ -0,0 +1,72 @@ | |||
| import abc | |||
There was a problem hiding this comment.
I believe this file can be removed
erikwijmans
reviewed
Nov 20, 2019
viewer.py
Outdated
| @@ -0,0 +1,87 @@ | |||
| import cv2 | |||
There was a problem hiding this comment.
Please remove this file from the PR
erikwijmans
reviewed
Nov 20, 2019
sensor_test.py
Outdated
| @@ -0,0 +1,61 @@ | |||
| import timeit | |||
erikwijmans
reviewed
Nov 20, 2019
There was a problem hiding this comment.
Can you add a test like https://github.com/facebookresearch/habitat-sim/blob/master/tests/test_sensors.py#L150 for the new noise models?
Deleting from PR.
Deleting from PR.
Removing from PR.
Merging origin/master to PR
|
Added noise models RGB outputs with default parameters to make noise mode choice easier. |
eundersander
pushed a commit
to eundersander/habitat-sim
that referenced
this pull request
Aug 6, 2020
Fix video generation
Ram81
pushed a commit
to Ram81/habitat-web-sim
that referenced
this pull request
Dec 10, 2020
Addition of Gaussian noise, Speckle noise, Salt and Pepper noise and Poisson noise for RGB sensors.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Motivation and Context
This change builds upon the redwood depth noise model implemented by Erik Wijmans with the objective of implementing color noise models in Habitat. This change includes initial commits for four new RGB noise models, namely Gaussian noise, Speckle noise, Salt and Pepper noise and Poisson noise. The hope for this pull request is to continue the efforts to allow for embodied agents to train under realistic sensor noise to further understand their behavior in sub-optimal sensor conditions.
Noise models RGB output with default parameters (added by @mathfac)
Gaussian Noise Model:
Poisson Noise Model:
Speckle Noise Model:
SaltAndPepper Noise Model:
How Has This Been Tested
These modification have been tested to gauge the frame rate at which the noise models are applied to arbitrary NumPy arrays as well as tested to see the visual effect of the noise models in the simulation environment. These tests can be replicated using sensor_test.py and viewer.py respectively.
Types of changes
Checklist