Called Seam Carving for Content-Aware Image Resizing, it's described as
Seam carving is a method for content-aware resizing that changes the size of an image according to its content. This paper shows applications to aspect-ratio change, image retargeting, and object removal.
Resizing images to make them fit in the required internet form factor without distorting its contents has always been a bit of a challenge. The demonstration of this tool, seems to make the work involved a lot easier and automated. The ability of a page to resize with the browser window is called liquidity. Here the images in the page also support that concept.
It also demonstrates the capability of locating areas of that are least important or have the 'least energy' and removing them without any noticeable distortions. However, there are limitations, especially where there are large composition elements such as faces and people that cannot be skewed without obvious effects.
To see these very interesting image transformation capabilities see the following video.
The two researchers, Shai Avidan and Ariel Shamir have been promptly recruited by Adobe Systems Incorporated, who in-fact chaired the conference session. I fear this tool may also just get incorporated into Adobe's proprietary image manipulation software, instead of becoming available in the open source arena.
Meanwhile, traditionalists moan the manipulations possible in digital images and note that it undermines the integrity of photographs and the art of photography.