A Video Is Not Just a Video
The evolution of video technologies has hyper-changed and hyper-charged the advertising and publishing industries. Multiple cycles of maturation and hype have occurred over the past two decades, and we are still only at the beginning of what’s actually possible. Video technologies are now revolutionizing how humans shop, entertain, and communicate. Simultaneously, all businesses are struggling to leverage the right solutions to adapt and evolve, lest they get disrupted by known and unknown competitors.
One of the more significant changes that is just now starting to escape the back rooms of universities and to touch our everyday lives via online content is in computer vision algorithms. Major companies are leveraging these algorithms to filter, distribute, search, monetize, and entertain.
Computer Vision and Our Future
There are several different terms to describe these technologies, from computer vision to image recognition, deep learning to artificial intelligence. While this all may sound quite futuristic, many of the core aspects are becoming a reality fast.
Computer vision “is a field that includes methods for acquiring, processing, analyzing, and understanding images and video. A theme in the development of this field has been to duplicate the abilities of human vision by electronically perceiving and understanding an image.”
A recent growing war among internet giants to hire and acquire computer vision and artificial intelligence talent is a testament to how fast and furiously these algorithms will impact billions of people, both on- and off-line.
Analyzing images and/or video to create more meaningful connections to audiences is at the heart of these acquisitions.
Some notable examples:
Google acquired Deep Mind for a reported a $500 million, and much ink has been spilled opining on what the Deep Mind’s team might now be working on. Users are uploading 100 hours of new video to YouTube every minute. Deep Mind could help “better curate the millions of videos on YouTube, making suggestions and related videos much smarter,” says Geoff Duncan of Digital Trends.
Dropbox acquired Anchovi, led by two computer vision PhDs, Boris Babenko and Peter Welinder. Serge Belongie and Pietro Perona, professors at Cornell and Caltech respectively, have also worked with Anchovi. At the time, they were working on artificial intelligence algorithms to help classify images, something that could surely empower Dropbox users — and the value of Dropbox’s IP.
“Computer Vision offers the potential to make users’ personal photo collections searchable based on image content. Face recognition is only the tip of the iceberg.” says Belongie, professor at Cornell NYC Tech with a focus in computer vision and machine learning.
Yahoo acquired image-recognition startup, IQ Engines, whose team has joined Flickr. Will this make future releases of Flickr easier for prosumers and hobbyists to search and find their photos? Will it better enable brands to find their logos and either pull them or monetize them?
There will be many more acquisitions of businesses and teams that specialize in computer vision. The bigger question is how does this impact brands, advertising, content distribution, and publishing.
Television historically was the place for brands to communicate their messages to the masses with the help of major ad agencies. At first there were only a couple of channels that charged a premium and then cable and satellite TV delivered hundreds of channels, which allowed almost anyone buy an ad for a specific market at different costs.
Now there are many more possibilities to market brands from Twitter or Facebook, to Google and Bing, to YouTube. Pre- and post-roll video, banners, email, overlays, and yes, still print and TV.
There’s a New Playground for Brands by Leveraging Computer Vision Tech
The news of early brands starting to test ad campaigns on Instagram videos is one more new social advertising opportunity.
It is fascinating that Instagram is being very selective with which advertisers it’s allowing to beta test the sponsored videos service. It makes sense as Instagram wants to maintain quality of the platform and while also trying not to scare away the loyal users.
Facebook “has also created a new research laboratory with the ambitious, long-term goal of bringing about major advances in artificial intelligence,” New York University professor Yann LeCun wrote. Therefore, it is possible that soon brands will be able to use their Instagram or Facebook advertising dashboard to place ads based on what’s inside the images and videos.
For example, maybe the Gap would like to advertise a 15-second pre-roll in front of an Instagram video that includes more than five multi-racial kids eating ice cream in a cup, with baseball caps on, a basketball court in the background. This is possible now but the technology needs to be more accurate in order to post advertisements appropriately.
Or, imagine if HBO was soon able to place ads for the next “Game of Thrones” episodes in front of any reference of any YouTube video that references the show (as long as the reference isn’t negative). Several companies are providing this service now but it’s mostly a process edited and filtered by humans rather than computer vision.
Very soon this all will be possible with the help of computer vision and artificial intelligence algorithms.
Our days at Cake-Works tend to be filled with the challenges of imaging technology; we find ourselves focused on fixing issues across tech, operations, strategy, and organization. We are tapped more often for problem-solving in a complex ecosystem instead of capturing and enjoying a new opportunity.
Sometimes, it’s important that all of us in the imaging and video businesses to stop grinding away at the nitty-gritty of the platforms and business cycles, and have a moment to be amazed. To watch an inspiring video all the way through. To appreciate the direction, light and artistry of a new web series. To remember that content will always be king, and that we got into this world because we love the actual art form, and not just the business around it.
Stop and smell the roses, and take a picture of them, and post it. And then look at five more from people you’ve never met, or maybe you have, knowing that it was an algorithm that surfaced such incredible content to you, and then tell those creators much they’ve inspired you.
Evan Nisselson invests in digital imaging and video technology companies via LDV Capital following a career of serial entrepreneurship, and mentoring and consulting on digital imaging products and platforms. You can find him at [email protected]
Rebecca Paoletti’s passion for film and video took her across startups, agencies and major media companies, in both product and business roles. Two years ago, having wrapped a several year stint running video at Yahoo, she co-founded Cake-Works, a digital video firm that supports both startups and Fortune 500 companies create and manage sustainable video initiatives. You can find her at [email protected]
Evan and Rebecca are also partnering with other imaging and video experts and over 30 distinguished speakers on an upcoming one-day imaging and video technology conference, the LDV Vision Summit.Tags: Artificial Intelligence, Computer Vision, Evan Nisselson, guest column, Imaging, LDV Capital, LDV Vision, Rebecca Paoletti, Startups, voices