Equalization operator #4575
Merged
Equalization operator #4575
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
stiepan
force-pushed
the
equalize_operator
branch
2 times, most recently
from
ae107e5
to
4551a07
Compare
mzient
reviewed
Jan 18, 2023
| assert(layout.size() == 2 || layout.size() == 3); | ||
| output.SetLayout(layout); | ||
| kernels::DynamicScratchpad scratchpad({}, AccessOrder(ws.stream())); | ||
| ctx_.gpu.stream = ws.stream(); |
There was a problem hiding this comment.
ctx should be a local variable - you populate all of its fields anyway.
mzient
reviewed
Jan 18, 2023
| imgs = [np.array(img) for img in imgs.as_cpu()] | ||
| assert len(equalized) == len(imgs) | ||
| baseline = [equalize_cv_baseline(img, layout) for img in imgs] | ||
| check_batch(equalized, baseline) |
There was a problem hiding this comment.
Do we expect bit-exactness with OpenCV?
There was a problem hiding this comment.
As discussed elsewhere, the algorithm is the same, the only float arithmetic is scaling of the comulative. It happens to work, let us see if the CI does not complain.
mzient
approved these changes
Jan 18, 2023
klecki
reviewed
Jan 18, 2023
| template <int ndim> | ||
| TensorListShape<2> GetFlattenedShape(TensorListShape<ndim> shape) { | ||
| if (shape.sample_dim() == 3) { // has_channels | ||
| return collapse_dims<2>(shape, {{0, shape.sample_dim() - 1}}); |
There was a problem hiding this comment.
The layout can be HWC and CHW, doesn't it impact anything here?
There was a problem hiding this comment.
SequenceOperator handles the latter
klecki
approved these changes
Jan 19, 2023
Signed-off-by: Kamil Tokarski <ktokarski@nvidia.com>
stiepan
force-pushed
the
equalize_operator
branch
from
4551a07
to
be9cb31
Compare
Signed-off-by: Kamil Tokarski <ktokarski@nvidia.com>
|
CI MESSAGE: [7059662]: BUILD STARTED |
|
CI MESSAGE: [7059662]: BUILD PASSED |
aderylo
pushed a commit
to zpp-dali-2022/DALI
that referenced
this pull request
Mar 17, 2023
Signed-off-by: Kamil Tokarski <ktokarski@nvidia.com>
Closed
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.
Signed-off-by: Kamil Tokarski ktokarski@nvidia.com
Category:
New feature (non-breaking change which adds functionality)
Description:
Adds per-channel histogram equalization operator (GPU, uint8_t).
Additional information:
Uses kernels added in the #4565
It is needed to implement the standard RandAugment/AutoAugment policies.
Affected modules and functionalities:
Key points relevant for the review:
Tests:
Checklist
Documentation
DALI team only
Requirements
REQ IDs: N/A
JIRA TASK: DALI-3187