add torch implementation for interpolation

[xycdx]

• replace cv2 bilinear interpolation with torch
• Benchmark shows 2x speedup compared to CPU version
• Closes FastBlendSmoother加速一倍的办法 #817

[gemini-code-assist]

Summary of Changes

Hello @xycdx, 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 transitions the interpolation logic within the PatchMatcher class from CPU-bound OpenCV operations to GPU-accelerated PyTorch functions. The primary goal is to enhance performance by utilizing torch.nn.functional.interpolate for both image resampling and Nearest Neighbor Field (NNF) updates, leading to faster processing times.

Highlights

• Migration to PyTorch for Interpolation: Replaced OpenCV's cv2.resize with torch.nn.functional.interpolate for image resampling and Nearest Neighbor Field (NNF) upscaling, leveraging PyTorch's GPU capabilities.
• Performance Improvement: This change is expected to provide a significant performance boost, with benchmarks indicating a 2x speedup compared to the previous CPU-based implementation.

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 in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

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 issue 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 replaces the CPU-bound cv2.resize operations with GPU-accelerated torch.nn.functional.interpolate for image and nearest-neighbor field resampling. This is a great change that should significantly improve performance by avoiding costly data transfers between GPU and CPU. While reviewing, I found a pre-existing critical bug in the update_nnf function that was made more visible by the formatting changes. The indexing logic for upscaling the nearest-neighbor field uses the wrong dimension size, which can lead to incorrect results or crashes. I've provided a correction and a more efficient implementation for that part.

[Artiprocher]

@xycdx 很强！感谢！