During its annual MAX user conference, Adobe routinely presents a variety of research initiatives that could potentially be integrated into its Creative Cloud suite of applications. A recently unveiled project, Project Sharp Shots, which will be highlighted later today at the MAX Sneaks event, is one I anticipate will soon be available within their video editing tools. Utilizing Adobe’s Sensei AI technology, Sharp Shots is a research endeavor focused on utilizing artificial intelligence to sharpen blurry videos.

Shubhi Gupta, the Adobe engineer leading the development of this project, explained that the core concept is to clarify videos—regardless of whether the blurriness stems from camera instability or rapid motion—with a single click. Demonstrations she provided showcased results ranging from slight improvements, such as in a video of her playing the ukulele, to significant enhancements, exemplified by a video of a quickly moving motorcycle.

“Unlike conventional techniques that require extensive parameter adjustments, Project Sharp Shots eliminates the need for that complexity,” she stated. “It’s a straightforward, one-click solution. It’s not based on trickery, but rather on the application of deep learning and AI, which analyze individual frames, reduce blur, and generate clear, high-resolution images and videos.”

Image Credits: AdobeGupta further explained that the team built upon existing research concerning image deblurring, adapting and refining the process for use with video footage, and then optimizing it for efficient memory usage and processing speed.

It is important to note that After Effects currently provides tools for deblurring and stabilizing footage, but these rely on a distinct algorithm with its own inherent constraints.

This new system performs optimally when it can analyze several frames both preceding and following the target frame, though it is capable of functioning effectively even with limited frame availability within a video.