Mind Boggling A.I. to Fix EXTREME Blur!

PiXimperfect
10 Dec 202021:45

TLDRIn this video, Umesh from Pixar shares his astonishment with a mobile AI application that can enhance and fix extreme blur in photos. He demonstrates the app's ability to bring details back into out-of-focus faces and restore old photos. Despite some limitations, such as a daily image limit and resolution restrictions, Umesh is impressed by the app's potential for both casual and professional use, suggesting that with further development, it could be a valuable tool in the world of photo editing.

Takeaways

  • 🤯 Umesh from Pixar discovers a mind-boggling AI mobile application that can fix extreme blur in photos, challenging the belief that desktop software is superior.
  • 🌟 The AI app can enhance an out-of-focus face in a photo by inventing details that were not originally captured.
  • 🔍 Before and after comparisons demonstrate the app's ability to bring back details in faces and even firecrackers.
  • 📸 The app uses a cloud-based system, possibly matching facial features with a database of sample images to fill in missing details.
  • 👨‍👩‍👧 The video includes examples of restoring old photos and improving low-resolution images, showing the AI's potential for professional work.
  • 🚀 The AI's performance is described as 'absolute magic' and 'totally out of this world', with impressive results in image enhancement.
  • 💻 The app has limitations, such as a daily limit of five free images and not exporting in the highest possible resolution.
  • 🎨 Practical applications of the app include fixing blurry faces in selfies and restoring focus in old or low-quality photos.
  • 🛠️ The video suggests using external upscalers to increase the resolution of the AI-enhanced images for professional use.
  • 💡 Umesh expresses a desire for the AI app to have a desktop version or plugins for Photoshop and Lightroom, highlighting the potential for professional adoption.
  • 💸 The app offers a free version with ads and a paid version without ads, prompting users to consider their needs and frequency of use when deciding whether to subscribe.

Q & A

  • What was Umesh's initial perception about mobile applications compared to desktop software programs?

    -Umesh initially believed that desktop software programs were superior and were the standard for professional workflows, often neglecting the capabilities of mobile applications.

  • How does the AI application work in enhancing out-of-focus images?

    -The AI application works by using cloud-based technology, where it presumably has access to a vast database of images. It matches the unclear parts of the image with the most similar ones from its database and adjusts them accordingly to fit the original image, effectively 'inventing' details where there were none.

  • What limitations did Umesh observe while using the AI application?

    -Umesh noted that the AI application had a limitation in exporting images at the full resolution it is capable of. It also inserted details like eyelashes and altered some facial features, which were not present in the original image.

  • How did Umesh test the AI application's ability to restore old photos?

    -Umesh tested the AI application's restoration ability by using an old photo of his parents graduating from art college. He wished to see more details on their faces, so he used the application to enhance the image and compared the before and after results.

  • What was the result of the AI application's attempt to restore a very low-resolution image?

    -The AI application managed to significantly improve a very low-resolution image (100 by 100 pixels). While it did not fully restore all details, it made a night and day difference, showing the potential of AI in image enhancement.

  • How did Umesh address the limitation of the AI application's output resolution?

    -Umesh used external upscalers like Gigapixel AI to increase the resolution of the images processed by the AI application. This allowed him to get a higher resolution output than what the AI application exported.

  • What practical applications does Umesh suggest for the AI application?

    -Umesh suggested practical applications such as restoring blurred faces in images and correcting focus失误 in photos. He demonstrated how the AI application could be used to bring back focus on faces and then composite the enhanced parts back into the original image using Photoshop.

  • How does the AI application perform in comparison to traditional methods of image enhancement?

    -The AI application performed remarkably well in comparison to traditional methods. It saved significant time and effort that would otherwise be spent manually painting in details or using other enhancement techniques. It was described as 'absolute magic' and 'totally out of this world'.

  • What is Umesh's recommendation for professionals using the AI application?

    -Umesh recommends that professionals can certainly incorporate the AI application into their workflow, but they should be aware of the current limitations, such as the export resolution and the need to transfer images between devices. He also suggests that the application could be even more beneficial if it allowed for higher resolution exports.

  • What are the pricing options for the AI application?

    -The AI application offers a free version that allows users to process up to five images per day. Users who need to process more images can choose to pay for the service to remove advertisements and potentially access higher resolution exports.

Outlines

00:00

🤖 Introduction to AI-Powered Image Enhancement

Umesh from Pixar introduces the audience to the revolutionary capabilities of a mobile application that challenges traditional desktop software programs. He expresses his surprise at the app's ability to enhance images, particularly focusing on the details that were previously impossible to recreate. The app demonstrates its power by taking a photo with a shallow depth of field and enhancing the out-of-focus face to bring it into sharp detail, a task that was unimaginable with conventional editing tools. Umesh is intrigued by the technology and plans to explore its potential further in the video.

05:00

🎨 Restoring Old Photos and Enhancing Resolution

The video continues with Umesh showcasing the app's ability to restore old photos and enhance their resolution. He tests the app on a decades-old photo of his parents, noting the app's impressive ability to add details to the faces. Despite some minor imperfections, such as slightly altered lip shapes and incorrect text on certificates, the overall improvement is remarkable. Umesh then pushes the app to its limits by testing it on a low-resolution image, amazed by the results as the app manages to add significant detail and improve the overall quality of the image.

10:01

🔍 Examining the Limitations and Potential of AI Enhancement

Umesh delves into the limitations of the app, noting that while it can enhance and restore images remarkably, it does not export in full resolution. He highlights the app's capability to process high-resolution images but points out that the exported results are limited in size and detail. Despite this, Umesh is impressed by the app's ability to recover details in blurred images and suggests that with further development, such AI technology could significantly impact professional workflows. He also discusses the potential applications of the app in various scenarios, including fixing out-of-focus shots and restoring old photographs.

15:01

💡 Practical Applications and Workflow Integration

Umesh explores the practical applications of the app, demonstrating how it can be used to bring focus back to blurred faces in a photo. He explains the process of enhancing an image on the app, exporting it, and then integrating it back into Photoshop for further adjustments. He also discusses the potential of the app in professional settings, suggesting that it could save significant time and effort for restorers and editors. Despite the current limitations, Umesh sees value in incorporating the app into one's workflow for specific tasks, and he provides suggestions for improving the app's functionality for professional use.

20:01

📢 Conclusion and Final Thoughts

In the conclusion of the video, Umesh reflects on the app's capabilities and potential impact on the industry. He expresses his hope for future updates that would allow for higher resolution exports and the possibility of a desktop version or plugins for popular editing software. Umesh acknowledges the app's current limitations but emphasizes its potential to save time and effort, especially for non-professional users. He wraps up by encouraging viewers to explore the app and consider its use in their own projects, and he thanks the supporters of his channel.

Mindmap

Keywords

💡EXTREME Blur

EXTREME Blur refers to an excessively blurred image where important details are significantly obscured, making it difficult to discern key features. In the video, the AI application 'Remedy' is demonstrated to correct such extreme blurring, restoring clarity and detail that were previously lost. This capability is shown through before-and-after comparisons of images that have been enhanced to reveal details like facial features and textures that were not visible in the blurred versions.

💡mobile applications

Mobile applications are software programs designed to run on smartphones, tablets, and other mobile devices. In the video, the narrator initially skeptical about the capabilities of mobile apps compared to desktop software, discovers a powerful mobile application for image enhancement that challenges his preconceptions. This application impressively corrects extremely blurred images, showcasing the advanced capabilities that mobile technology can offer.

💡enhance

In the context of the video, 'enhance' refers to the feature of the mobile application that improves the quality of an image by automatically correcting blurs and refining details. The narrator demonstrates this feature by selecting a blurred image, applying the 'enhance' function, and showing the significantly improved clarity in the processed image, effectively bringing obscured details into sharp focus.

💡sample images

Sample images are likely reference photos used by the AI application to reconstruct blurred or missing details in the images being enhanced. The narrator speculates that the application uses sample images to match and replace features like eyes, lips, and facial contours in blurred photos, thus inventing details that appear in the processed image but were not discernible in the original.

💡cloud-based software

Cloud-based software refers to applications that run on remote servers rather than on a user's local computer or device. The narrator mentions that the image enhancing app operates in the cloud, suggesting that heavy processing tasks, such as image analysis and detail enhancement, are handled by servers. This allows the application to utilize more powerful computing resources than those typically available on mobile devices.

💡restoration

Restoration, in the context of the video, involves repairing and improving old or damaged photos using the AI application. The narrator demonstrates this by enhancing an old graduation photo to bring out more facial details and improve overall image clarity, illustrating how AI technology can be used to preserve and restore visual memories.

💡resolution

Resolution in digital imaging refers to the number of pixels contained in an image, which affects its clarity, detail, and the sharpness of its rendering. The video discusses various tests on images with different resolutions to see how well the AI application can enhance low-resolution images to make them clearer, as seen when a 0.01 megapixel image is enhanced to reveal much more detail.

💡upscalers

Upscalers are tools or software that increase the resolution of an image, making it larger and often clearer without losing significant quality. The narrator explores using external upscalers like Gigapixel AI to further enhance the resolution of images processed by the mobile app, highlighting how combining different technologies can improve the usability of restored images in professional contexts.

💡export limitations

Export limitations refer to restrictions on the resolution or quality of images saved from the application. The video discusses how the AI application does not allow images to be exported at their highest possible resolution, which could hinder professional use unless additional steps are taken, such as using external upscalers to achieve desired resolutions.

💡professional workflow

Professional workflow in the context of the video refers to the processes and tools used by professionals to achieve high-quality results in their work. The narrator evaluates whether the mobile app's capabilities can be integrated into professional workflows for tasks like unblurring and restoring images, discussing the challenges and benefits of using mobile applications in professional settings.

Highlights

Umesh from Pixar shares his amazement with a mobile application that outperforms desktop software in fixing extreme blur in photos.

The application can focus on details such as faces even in a photo taken with shallow depth of field where the face was originally out of focus.

The AI-based software seems to invent details where there were none, improving the quality of the image remarkably.

The enhancement process is fully automatic, requiring only a click on the 'enhance' button and a tick to complete.

Before and after comparisons showcase the software's ability to bring back details in faces and even firecrackers.

The software's effectiveness is so impressive that Umesh is not opposed to it advertising itself as an AI software.

The video will explore the application's potential for professional use and its limits.

The application may use sample images to match and adjust details in the original photo, suggesting a cloud-based processing mechanism.

In the restoration of an old photo, the software manages to add details like teeth and spectacles, although it sometimes changes original features.

The software can recover a low-resolution image, but it does not export in full resolution, which is a limitation for professional use.

An extreme test shows that the software can recover details from an image with only 0.01 megapixels.

The software's ability to recover and add details is so impressive that Umesh would be the first to purchase a Photoshop plugin if available.

The software can recover details in a slightly blurred image, retaining much of the original eye details.

For professional use, the software's challenge is the workflow involving transferring images between computer and phone and the limitation on resolution.

External upscalers can be used to increase the resolution of the exported images, overcoming the software's limitation.

Practical applications include restoring focus to blurred faces and creating a shallow depth of field effect.

The software is free to use with a limit of five images per day, which is sufficient for most users.

Umesh suggests that the software could make a lot of money if it offered higher resolution exports for professionals.

The video concludes with a recommendation to try the software, especially for those who unblur or restore portraits, as it can save a significant amount of time.