Effective video communication is important for everyone who communicates online. For businesses, educators, and content creators, it is vital.
NVIDIA Maxine is a suite of NVIDIA-accelerated SDKs, cloud-native containerized NVIDIA NIM microservices for deploying AI features that enhance real-time audio and video for video conferencing, digital humans, virtual presence, and content creation. Maxine continues to empower you to push the boundaries of what’s possible in intelligent real-time video enhancement.
In this post, we showcase how NVIDIA Inception program partner VideoRequest uses the Maxine AI developer platform to bring sophisticated, real-time, video-editing tools to their platform, enabling end users to focus on the message they want to send and not the video settings or processing tools.
Transforming video creation with AI
VideoRequest is an AI-enhanced video service where users can easily create, edit, and share videos for testimonials, advocacy, marketing, training, or events. The company’s built-in video editor and integrations to social media provide users with asynchronous video communication in a single easy-to-use service. VideoRequest’s customers include associations, nonprofits, and higher education groups that rely on authentic communication and engagement.
Simplifying video creation and editing
Maintaining eye contact is a critical aspect of increasing authenticity, to create more engagement and deliver a polished result. Presenters looking at a script or focusing away from the camera appear less connected with their audience. Distracting backgrounds are just that—distracting!
Maxine enabled VideoRequest to solve these problems for users with state-of-the-art AI models and features:
- Maxine Eye Contact
- Maxine Virtual Background
Maxine Eye Contact
This feature uses AI to redirect the user’s gaze so that they appear to be looking directly at the camera in a natural way. Direct eye contact creates a more engaging and personal connection with viewers.
Eye contact is an optional feature within VideoRequest, but since integration, it’s become VideoRequest’s most popular feature.
“Integration with NVIDIA Maxine enables cutting-edge AI enhancements with an intuitive user experience. Most of our users look at themselves on the screen, or a script, rather than the camera. Eye Contact instantly warms up their video submissions and makes them more personal. This is especially important for impact statements, testimonials, sales pitches, and appeals videos,” said Terry Nawrot, CEO of VideoRequest.
Maxine Virtual Background
Users can easily replace their original background with a virtual one, providing a professional look regardless of their actual physical environment.
“VideoRequest applies the background replacement filter to every video that our users create. This way it makes it quick for users to swap out backgrounds whenever they choose,” Terry Nawrot explained.
According to Nawrot, their previous system would take 9x the video duration for processing time to do background replacement with other software. A 1-minute video would take 9 minutes of processing time. With the NVIDIA real-time solution, the same task takes about 0.8x the duration, so a 1-minute video takes under 50 seconds of processing time using the NVIDIA Triton release on Linux.
Through the 10-month partnership with NVIDIA, the VideoRequest team has seen a massive speed improvement, as well as quality enhancements in other features adopted from NVIDIA.
Real-world impact
The integration of NVIDIA Maxine into VideoRequest’s platform has already shown significant benefits for users across various industries, including online communities and marketing.
Communities require strong, lasting relationships with their members. However, in the absence of frequent in-person events and interactions, they rely on video communication to engage, attract, and retain their members.
VideoRequest’s partners see a lot of value in using Eye Contact, which elevates engagement and authenticity in the videos, and therefore increases retention. It elevates the quality of their testimonial videos to a level they can use in marketing, despite having been created by regular people who normally look away from the camera to read a script.
Beth Arritt, president and CEO of The Arritt Group, found that using Maxine’s Eye Contact in VideoRequest significantly enhanced customer testimonials. “With eye contact, the video now exudes confidence and authenticity.”
In the following video, you can see a before-and-after version of a testimonial.
The final video output has both Eye Contact and Virtual Background for a polished result.
Looking ahead
From enhancing day-to-day video conferencing to integrating AI technology, NVIDIA Maxine offers high-quality video communications for all professionals.
The latest Maxine production release is included exclusively with NVIDIA AI Enterprise, which enables you to tap into production-ready features such as NVIDIA Triton Inference Server, enterprise support, and more.
As NVIDIA continues to develop and refine Maxine’s real-time video and audio AI features, we’re excited to see startup partners like VideoRequest simplify high-quality video creation, making it accessible to businesses and individuals alike.
With the growth in content creation and the need for video in marketing, communication, education, and business, the partnership between NVIDIA and VideoRequest is poised to drive innovation and improve video communication experiences for users, especially as the VideoRequest team is working on integrating more Maxine features soon.
For more information about Maxine and how it can enhance your video applications, see the NVIDIA Maxine developer page. If you’re interested in early access, with non-production access to production and soon-to-be-released features, see the Maxine Early Access program (requires login).
For more information about the VideoRequest AI-enhanced video creation platform, see VideoRequest. For more information about Maxine features, see Advancing Telepresence and Next-Generation Digital Humans with NVIDIA Maxine and get a preview of the Eye Contact NIM microservice in the NVIDIA API Catalog.
To help improve features in upcoming releases, provide feedback with the NVIDIA Maxine and NVIDIA Broadcast App survey.
