Monday, 2 March 2020

AI is changing the way we edit videos

Redshark News
AI tools for automating video and finishing posting and posting of polished video clips on social media just took two giant leaps forward with the launch of Vimeo Create and fresh multi-million dollar funding for start-up Revl. Both are aimed initially at the burgeoning market for time-strapped but video hungry business marketers.
Vimeo Create is a short-form video editing platform built out of Vimeo’s acquisition last year of Magisto for a reported U$200 million. Backed by Qualcomm, Magisto analyses images, video, speech and audio uploaded via its app to automate the production of a suitably professional clip for sharing on social. Users can select from a gallery of professionally designed video templates, or they can create a video from scratch using their own footage and storyboard.
“It’s a radically simple tool that shortens the distance from idea to execution so more businesses can have a successful video strategy,” said Vimeo chief-executive Anjali Sud. 
Venture Beat notes that Magisto “could play a big role in Vimeo’s broader pivot away from its former ‘YouTube alternative’ status as it looks to position itself as a place for creatives and businesses to access the tools they need to make videos.”
Video is vital for every business, yet the barriers to creating video remain high. Vimeo's own research found that while over half of small businesses used video in their marketing strategies last year, only 22% feel that they're using enough video.
“The research is clear: small business owners and entrepreneurs don't have the tools, time or budgets to make videos at the volume and quality needed to compete," added Sud.

The wider implications

San Francisco-based Revl just raised $5.2 million in a funding round led by Nimble Ventures, Silicon Valley Data Capital and Luma Pictures. That brings the five-year-old Revl’s total funds to $10 million.
The funding will be used to fuel the growth of its AI video production service, Revl X with which it hopes to encourage customers of ‘adventure experiences’ to upload videos and share online as a form of word-of-mouth marketing.
Co-founder Eric Sanchez admits, in a press release, that editing and delivering fully customised ‘video souvenirs’ of epic experiences in real-time is computationally and programmatically very difficult. Revl has apparently used AI and parallel processing to build a fully automatic system that edits and delivers HD cinematic videos in seconds.
What is unique about Revl is that this only works with its own hardware – an action camera which would be loaned to customers by the operators of adventure experiences – think skydiving, racing/driving, alpine coasters, and zip-lines.
How it works is that users select a video package from an Onboarding app and are provided with a QR sticker, which the 4K Arc scans at the beginning of each video. Cameras docked into an Editing Box module on-site which automatically transfers and analyses footage, wipes the SD card, charges the camera and edits the videos and photos, complete with B-roll and animated intros and outros, with RAW files sent to Revl’s cloud.  
“The user will find the finished video in the Revl app and can modify the music and share directly to social media,” it explains.
The Arc has a 12 Megapixel sensor with 150-degree field-of-view lens and will capture 4K video 30fps. It has a gimbal integrated into the rear of the camera which keeps it level to the horizon. It also has electronic image stabilisation which reduces bumps and camera shake. It’s waterproof up to 33ft (10m) and houses a 128GB MicroSD memory card plus two rechargeable lithium-ion batteries rated at 800mAH, 3.7V, 2.96Wh.
Revl also claims its solutions “understand human emotions” to identify key moments in an experience and that the cameras are “geospatially aware”. The AI uses this information to auto-create the clip.
The content also comes with hashtags and other key identifiers pre-embedded, so brands are automatically included in social posts.
“Our goal is to make our technology vertical agnostic in the next 1-2 years so it can be used in almost any kind of adventure or activity,” Sanchez said. “Just like you would expect on the Splash Mountain ride but with a video instead of just a photo, for example.”

No comments:

Post a Comment