In the synthetic media world, image and video synthesis use artificial intelligence (AI) to generate media content. The technologies can be used to create human avatars as well as digital worlds without the need for actors or physical sets, providing the potential to create photo-realistic visual effects that could transform gaming, movie production, enterprise business video content creation, and more.
As part of our Synthetic Media Landscape, Samsung Next identified five leading companies in the image and video synthesis category. All of these companies are developing AI-based solutions to automate the creation of video and image based content.
Lalaland uses neural networks to generate images of human avatars for eCommerce apparel retailers. It enables brands to customize online shopping experiences based on consumer age, size, and ethnicity. The platform enables online shoppers to see themselves in the latest apparel using personalized models size. Users can adjust their online display avatar in real-time to more closely resemble themselves wearing a particular outfit or article of clothing.
B2B, B2C, Women’s Fashion, eCommerce, Resale, Digital Fashion, Manufacturing
RefaceAI uses generative adversarial networks to enable face-swapping in videos and photos within the Reface and Reflect application creator -- interfaces which were developed by the company. Its technology can perform exchanges from 2D photos to 3D video with the use of one photograph and its API facilitates real-time development of content. The platforms provided by RefaceAI enable users to develop new internet content like memes and videos, allowing users to exchange the face of a celebrity with their own or vice versa.
B2B, B2C, Video Production, Content Creation, Marketing
Rephrase is building a generative AI platform to ease video creation, that can take a piece of text and create a video of a person speaking that text. A deep learning engine lies at the heart of Rephrase that learns facial features corresponding to spoken audio, and reconstructs photorealistic faces for any new given text or audio. This engine can be accessed by a simple API call or a web app.
B2B, Enterprise, Corporate Training, Marketing, Customer Service
RosebudAI enables the algorithmic placement of any face onto any body in their collection of stock photos. It is their first step in their synthetic stock photo and API offering, which will eventually allow users to edit and fully synthesize visual content with an intuitive interface.The company has also developed a product called “Tokking Heads,” a text-to-animation product that enables users to generate a human-realistic or virtual avatar using only text and photo. The cloud-based application also facilitates real-time editing and customization.
B2B, B2C, Entertainment, Fashion, Marketing
Synthesia is an AI solution enabling video and content creation without the use of cameras, actors, or physical studios. Its flagship program, CREATE, helps users produce content in just a few minutes using in-house or user created AI presenters.The script is capable of being translated into over 38 different languages. The system is well-suited for presenter-led video courses and corporate communications by enabling users to generate human-like AI avatars with an API infrastructure that allows real-time generation of content.
B2B, B2C, Video Production, Corporate Communications, Education, Sales
What is all means
Image and video synthesis are some of the most immediate applications of synthetic media that consumers will be exposed to. Whether it be through advertisements, movie production, training videos, or simple fun memes, it is already beginning to change the way media is produced and consumed. For the most part, it looks like AI will help democratize creativity to a wider pool of creators by allowing for high quality content on lower budgets.
To learn more visit http://syntheticmedialandscape.com