AI video creation — always a fascination
3 innovation and digital news in 1 minute. Every Monday. Episode 365
Let AI-avatars read any text with Synthesia
Synthesia can create videos with AI voices and AI avatars in more than 120 languages. The core technology uses deep learning algorithms to synthesize realistic facial expressions. Synthesia does research on synthesizing human motion and expression with avatars. Expanding from facial expression to full-body motion with their latest research paper on their concept “HumanRF”. In a video of a moving person, “HumanRF” simulates perspectives and camera angles that were not captured in the original. Opinion: At Space and Lemon Innovations, we often used Synthesia for quick explanation videos or as a funny opener for presentations. Features such as a music library, captions, and animations make Synthesia extremely useful. We presented Synthesia for the first time in 2021, back then it was more of a “WOW” reaction from clients.
AI-avatar for short explainer videos, great!
Translate video into any language with HeyGen
HeyGen is an AI-powered video generation platform. Use own video or an AI avatar. Key feature is translation of videos. Translates into a variety of languages while keeping the speaker voice and making the mouth move accordingly. Two weeks ago launched “Avatar 2.0, instant avatar”, generates custom avatar based on uploaded video. This video can be smartphone video, no professional greenscreen etc, needed. Custom avatar created in 5min (they call it “Instant avatar”). Opinion: HeyGen first amazed TikTok creators who then made videos in different languages with translation and went viral. The app finds a lot of use, not only for TikTok video creation, but also for businesses that need to cover multiple languages. Translation makes any kind of content more accessible in a matter of seconds.
Will HeyGen also establish virality outside of TikTok?
Photo to animated speaker video in seconds with D-ID
D-ID only needs one picture and a script to transform it into a talking-head video. Upload a picture of a face and the AI will animate facial movement that fits the script within seconds. It can composite faces and speech in 119 languages based on prompts from users’ descriptions. Opinion: The fascination is, it only takes one static image of a face for D-ID to make it into an animated video. Within 90 seconds. Very beginner friendly, the image doesn’t even need a green-screen. GPT integration can write a script draft.
D-ID started 2017, a very early Generative AI startup.
