I’m going to go out on an enormous limb here and claim he just made a very poor investment decision by placing a reportedly “upper five figures” bet on this startup.

ZunaVision was born out of the artificial intelligence lab at Stanford University by three researchers and an Assistant Professor, and what they’ve come up with is technology that enables video publishers – from amateurs who produce content for fun to professionals – to insert image and video advertising units into clips. It is product placment for online video.

This is how it works: an algorithm first analyzes the video, subsequently alters different aspects of embedded images or videos (such as the lighting, color and texture), and then attempts to fit the advertising into the physical space of their videos without appearing like a blatant overlay. Watch the video below for some example results.

If all this sounds familiar, it’s because it’s not an entirely new idea; I’ve covered a similar initiative called SmartAd from a company called Keystream in the past, and we’ve recently reviewed an Israeli company called Innovid doing similar things with in-video advertising.

My biggest gripe with these companies is not related to the technology, which is actually quite impressive, but the fact that these startups keep throwing words like ‘non-intrusive’, ‘uninterrupted’ and ‘natural’ at us while at the same time promoting the embedding technology to advertisers as ‘unmissable’ and ‘unskippable’. I find it unbelievable that they’re pretending to create a beneficial situation for publishers, advertisers and viewers while it’s painfully clear the latter are not gaining anything from this but a poorer viewing experience.

Online video is still notoriously hard to monetize, but I think ZunaVision is taking a step backwards by essentially giving everyone a way to clutter videos with product placement that is the video equivalent of pop-up ads

Crunch Network: CrunchGear drool over the sexiest new gadgets and hardware.

A lot of experimentation with video advertising on the Web is going on these days. There are interactive overlays, contextual text ads, video-in-video, ads that take over the whole screen, and little icon bugs that crawl along the bottom. Nevertheless, the plain video pre-roll that mimics a traditional TV commercial remains the most popular, even though everyone knows it sucks.

Today, an Israeli startup called Innovid (see earlier coverage) is doing its part to help advertisers move away from the pre-roll with technology that allows for things like virtual product placements inside videos and turning entire surfaces into clickable canvases. In the demo video above, for instance, you’ll see different advertising effects projected onto the walls of a room, from a Pong game that you can play to a RSS feeds running across the baseboard as a ticker to funny creatures whose eyes you can move with your mouse. (The video above is not interactive, but you can check out the interactive features at Innovid’s homepage).

Innovid also has an ad gallery that shows other types of clickable ads its technology can create, such as interactive product placements, video billboards (basically another video running in a frame on a wall), bugs, overlays, and hotspots. Frankly, I don’t find these quite as exciting as the clickable canvas approach. The product placements just look cheezy and unconvincing, at least in the old 1950s video Innovid chose to demo the feature. But turning an entire surface inside a video into a 3-D interactive canvas could be highly engaging if done properly.

This would definitely work for pure video ads that encourage interaction. It would be harder to pull off for a Web video show, but would be possible if a video was shot with enough blank spaces where ads could be inserted.

The technology could also be used eventually as a way to add an interactive layer to any video. However, the company is focusing on advertising opportunities first. Innovid offers both the ad server to place these ads and a dashboard to allow advertisers to keep track of mouseovers, clicks, and every other interaction. Innovid essentially places a frame over the original video where the interactive ad resides, complete with metadata. That extra layer makes the ad slots reusable and interchangeable, so different ad campaigns can be run in the same video over time, just like on a Webpage.

Innovid raised $3 million last year from Genesis Partners (jeff Pulver is also an investor).

Crunch Network: MobileCrunch Mobile Gadgets and Applications, Delivered Daily.

I had the chance today to chat with Zvika Netter, the CEO and co-founder of a new Israeli startup called Innovid that is working on technology that will empower video content producers to place interactive virtual items into their videos.

Innovid has remained in stealth mode up to this point and has only begun discussing its work now that it has raised $3M from Genesis Partners in a Series A round (previous angel funding came from Jeff Pulver).

The virtual items facilitated by Innovid are basically 3D objects that producers insert into videos post-production. They’re intended to look as realistic as possible so that they blend in with the real physical environment recorded by the video. And yet, they can’t go entirely unnoticed because users are encouraged to click and perform mouse gestures with them to derive additional functionality (for example, to view a popup description about the particular item with links to external resources).

The most obvious use of these items would be to simply drop branded items into user generated content. If you’re looking to monetize your video, you could, for example, sign a deal with a beverage company and place their product on the table during an indoors scene. The virtual item representing the beverage could then respond to camera movements; when the camera moves around to the left, you also see the left side of the beverage appear. This is possible because the 3D object has been mapped to its calculated surroundings.

But with Innovid’s object placement, you could also make it so that users who click on the beverage see a description of it and the stores in which it’s sold. Or you could allow the user to even move the beverage to another location within the video or have it perform a special effect when clicked on. Whatever the complexity, the object becomes a more effective advertisement through its interactivity.

Since these are virtual objects, they can also be served up differently depending on the intended audience. Certain countries or languages could see their own types of beverages. And of course, things could be tailored to the individual as well depending on their personal preferences.

Netter isn’t saying exactly when the company plans to debut its product or even show a demo to the public, but I had a chance to watch a video of these virtual objects and I must say it looks very compelling even at this early stage. The 3D items can be made to look realistic and their integration into video is quite seamless. This is a company that should be watched as the online video industry figures out how to monetize its content most effectively.

Crunch Network: CrunchBase the free database of technology companies, people, and investors

The general public is invited to parts of the show. more information is available on this website, including speakers.

Participating startups include 5min, Plymedia, AllofMe, Nuconomy, ClickTale, blogTV, Sportingo, PicScout, Qoof, 8hands, Velingo, Innovid, Semingo, PageOnce and Journeys.

I moderated a panel at the event last year. Highlights from that panel and the event in general are in the video below.

Crunch Network: CrunchBoard because it’s time for you to find a new Job2.0