Fix gaussian_blur padding calculation for even kernel sizes

[SamareshSingh]

Fixes #22046

Description

This PR fixes a bug in the NumPy backend's gaussian_blur function where it crashes when processing images with even-sized kernel dimensions.

The gaussian_blur function was using incorrect symmetric padding that caused shape mismatches when convolving with even-sized kernels.

[gemini-code-assist]

Summary of Changes

Hello @SamareshSingh, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request resolves a critical bug in the NumPy backend's gaussian_blur function that led to crashes when processing images with even-sized convolution kernels. The fix involves a precise adjustment to the padding logic, ensuring that the np.pad function correctly applies asymmetric padding when necessary, thereby preventing shape mismatches during the convolution operation.

Highlights

• Padding Calculation Fix: Corrected the padding calculation in the gaussian_blur function for the NumPy backend to properly handle even-sized kernels, preventing crashes due to shape mismatches during convolution.
• Asymmetric Padding Implementation: Introduced explicit asymmetric padding (pad_h_after, pad_w_after) to replace the previous symmetric padding, which was incorrect for even kernel dimensions and led to erroneous behavior.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request addresses a bug in the gaussian_blur function within the NumPy backend, which occurred when using even-sized kernels. The previous symmetric padding logic caused shape mismatches, and this PR corrects it by implementing proper asymmetric padding. This ensures that the output dimensions match the input dimensions after convolution. The fix is clear, correct, and well-targeted.

On a related note, the PyTorch backend's implementation of gaussian_blur in keras/src/backend/torch/image.py appears to have a similar vulnerability. The padding is calculated as int(kernel_size[0] // 2), which may not handle even-sized or non-square kernels correctly. It would be beneficial to investigate this in a separate issue to ensure consistent behavior across all backends.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 82.80%. Comparing base (b0e1dde) to head (14deba1).

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #22054   +/-   ##
=======================================
  Coverage   82.80%   82.80%           
=======================================
  Files         592      592           
  Lines       62340    62342    +2     
  Branches     9759     9759           
=======================================
+ Hits        51623    51625    +2     
  Misses       8193     8193           
  Partials     2524     2524           

Flag	Coverage Δ
keras	82.63% <100.00%> (+<0.01%)	⬆️
keras-jax	62.42% <0.00%> (-0.01%)	⬇️
keras-numpy	56.56% <100.00%> (+<0.01%)	⬆️
keras-openvino	37.57% <0.00%> (-0.01%)	⬇️
keras-tensorflow	63.67% <0.00%> (-0.01%)	⬇️
keras-torch	62.44% <0.00%> (-0.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[hertschuh]

Thank you for the fix!

Can you add a unit test that would fail without the fix? (I guess a test with an even kernel size)

You'll probably have to fix this one too:

keras/keras/src/ops/image_test.py

Lines 743 to 744 in b0de8d9

	(kernel_size[0] // 2, kernel_size[0] // 2),
	(kernel_size[1] // 2, kernel_size[1] // 2),