How AI is Tackling Marketers' Top Video Management Challenges

Last updated: 06-14-2020

Read original article here

How AI is Tackling Marketers' Top Video Management Challenges

There’s no getting around it — video has become a critical component of a marketer’s content toolbox. With 55% of people worldwide watching online videos every day, it’s an opportunity that no marketer can pass up. Short video formats, in particular, have taken off thanks to visual-focused social media platforms, like Instagram and Tik Tok. In fact, according to G2Crowd, social videos generate 1200% more shares than text and image posts combined.

But, while video is incredibly valuable, it can be a beast for marketing teams to manage — especially when working with terabytes of footage. Even when stored in a central location, like a digital asset management system (DAM), the inability to pin-point specific moments can still make managing videos a daunting task.

This is why the advancements being made with artificial intelligence for video are so critical. Leaders in the AI space, like Microsoft, are using AI to tackle worldwide video challenges, including accessibility, discoverability and monetization. For marketers, these additions can represent significant time and productivity savings.

Here’s a closer look at some of the awesome video advancements that are changing marketer’s lives for the better.

Object recognition is arguably one of the most prominent applications of artificial intelligence for marketers. When working with a ton of images, being able to instantly recognize simple characteristics like trees, a cat, etc., dramatically improves discoverability for high-volume teams. The challenge when you start tagging videos, however, is that it often isn’t enough to know that there’s a cat somewhere in the video. To truly drive efficiency, marketers need to be able to identify the exact moments in a video where the cat is present.

New artificial intelligence tools, like those available in Microsoft’s Audio/Video Indexer, not only analyze what objects are in the video, but where they are in the video. In addition to recognizing objects, this type of AI engine can recognize text, celebrities, common landmarks, key topics and more. When these capabilities are combined with a digital asset management system, they can save marketers countless hours, allowing them to instantly find the exact moment they need from any given piece of footage.

Katherine Tsai, Computer Vision Project Manager at Microsoft, shared insights on MediaValet’s webinar: “We’re really excited to see that AI is helping customers to grow efficiency and effectiveness, especially in [this] space. Microsoft has been investing significantly in engineering resources to develop best-of-breed AI capabilities to enable digital asset management.” 

The use of closed caption has increased significantly with the surge in videos on social media and as companies make themselves more accessible. In fact, a survey from 3Play Media found that 98% of survey participants frequently watch videos with the subtitles on. However, providing transcriptions for marketing videos can take hours. Even with a professional transcriptionist, it takes about 2-3 hours to transcribe an hour of video and can quickly become a costly expense. What’s more, once a transcript is created, it rarely provides time stamps that allow you to understand when any given phrase was said.

When transcripts are provided by an AI model, both time and costs are reduced considerably. In addition to providing accurate transcription of the text, these models provide time stamps and associate the text with each speaker. This enables marketers to instantly locate videos containing specific words spoken by a specific person. Some video AI tools can even translate video transcriptions into multiple languages.

From social media to advertisements, marketers are creating a ton of video. So much so, that 80% of the world’s internet traffic is expected to be video by 2021, according to recent data from Cisco. The key challenge that marketers face is how to create these videos at scale. Turning full-feature films into digestible, social-friendly clips can be a time-consuming process — especially when working with a creative team. 

Artificial intelligence can assist marketers with turning footage into short, social media friendly videos. AI-based editors will take a full-length film and break it down into digestible, logical clips that marketers can then select from to create an entirely new video to be shared. These capabilities allow marketers to quickly make trailers, social media posts, highlight reels and more.

Tsai continued on the webinar: “We’re optimizing existing general models and working with leading partners, like MediaValet, to deliver more powerful services to enable the indexing and searching of images, videos and documents based on what customers need.” 

Video marketing is just one of the many areas that artificial intelligence can help marketing teams boost productivity and enhance their digital marketing. Download your copy of 8 Ways AI Can Power Your Digital Strategy to see how you can get started today.


Read the rest of this article here