[✨ Enhanced Sample] Implement Generative Expand (OutPainting) with Imagen 3.0 #113
Conversation
madebymozart
requested review from
lethargicpanda,
calren and
ksemenova
as code owners
Summary of ChangesHello @madebymozart, 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 introduces the "Generative Expand" feature, which empowers users to creatively extend the boundaries of existing images using Imagen 3.0's outpainting capabilities. The core objective is to transform standard-sized generations into larger, more comprehensive scenes by intelligently filling in new areas while maintaining contextual consistency. This functionality is supported by backend integration with Imagen 3.0 and includes resolution safeguards to ensure optimal performance and user experience. Highlights
Using Gemini Code AssistThe 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
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 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
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request successfully introduces the Generative Expand (outpainting) feature using Imagen 3.0. The implementation is well-structured and integrates cleanly with the existing architecture. My review focuses on improving the robustness and maintainability of the new functionality. I've identified a potential user experience issue with the validation logic for the 'Expand' button, and I've also provided suggestions to address magic numbers, hardcoded strings, and minor stylistic improvements to align with best practices.
| onClick = { | ||
| onOutpaintClick(promptTextField) | ||
| }, | ||
| enabled = enabled && bitmap.width < 4096 && bitmap.height < 4096, |
There was a problem hiding this comment.
This enabled check is not robust enough. It allows the user to click 'Expand' on an image that will become too large after the operation (which doubles the dimensions), leading to a failed request and a poor user experience.
The validation should be predictive, ensuring the resulting dimensions are valid. For example, it should check if bitmap.width * 2 < 4096.
Additionally, the 4096 limit is a magic number and should be a shared constant.
| enabled = enabled && bitmap.width < 4096 && bitmap.height < 4096, | |
| enabled = enabled && (bitmap.width * 2) < 4096 && (bitmap.height * 2) < 4096, |
| } | ||
|
|
||
| fun outpaintImage(sourceImage: Bitmap, dimensions: Dimensions, prompt: String) { | ||
| if (dimensions.width > 4096 || dimensions.height > 4096) { |
There was a problem hiding this comment.
The value 4096 is a magic number. It should be defined as a constant (e.g., in a companion object) to improve readability and maintainability. This constant could then be shared with the UI layer (ImagenEditingGenerationInput.kt) to keep the validation logic consistent.
| fun outpaintImage(sourceImage: Bitmap, dimensions: Dimensions, prompt: String) { | ||
| if (dimensions.width > 4096 || dimensions.height > 4096) { | ||
| _uiState.value = ImagenEditingUIState.Error("Dimensions are too large. Maximum dimensions are 4096x4096") | ||
| return; |
There was a problem hiding this comment.
The semicolon in this return statement is not idiomatic in Kotlin and can be safely removed.
| return; | |
| return |
| contentDescription = "Outpainted image based on prompt: $prompt", | ||
| ) | ||
| } catch (e: Exception) { | ||
| _uiState.value = ImagenEditingUIState.Error(e.localizedMessage ?: "An unknown error occurred during outpainting") |
There was a problem hiding this comment.
This fallback error message is hardcoded. It's recommended to move user-facing strings to res/values/strings.xml to support localization.
Additionally, for better debugging, consider logging the full exception e before setting the error state, as e.localizedMessage can sometimes be null or lack detail.
madebymozart
changed the title
|
Let's wait until the redesign is done before we add new samples. |
Description
This pull request introduces the Generative Expand feature, a powerful implementation of outpainting that leverages the advanced capabilities of the Imagen 3.0 model. This new functionality empowers users to creatively extend the canvas of a generated image, seamlessly adding new content that is contextually consistent with the original.
The core objective is to allow users to transform a standard-sized generation into a larger, more comprehensive scene—whether it's widening a landscape, adding more context to a character portrait, or simply exploring what lies beyond the original frame.
Changes Introduced
Technical Details & Implementation
The outpainting process works by sending the original image along with the desired new canvas dimensions to the model. The original image acts as a contextual prompt, and Imagen 3.0 generates the pixels for the new, empty regions.
A key implementation detail is the introduction of a resolution cap.
4096x4096pixels.How to Test 🧪
3000x3000).4096x4096.Screenshots
| Before Expansion | After Expansion |