Home / Technology / AI Avatars Get Video Upgrade With New Diffusion Model
AI Avatars Get Video Upgrade With New Diffusion Model
23 Dec
Summary
- New diffusion model generates video avatars from single images.
- Lemon Slice secured $10.5 million in seed funding.
- Model supports live-streaming at 20 frames per second.
Lemon Slice has introduced its advanced diffusion model, Lemon Slice-2, poised to revolutionize AI-driven interactions by adding a video layer to chatbots and agents. This new technology can generate digital avatars from a single image, enabling them to perform various roles, including customer service, academic assistance, and mental health support.
The startup, founded in 2024, has successfully secured $10.5 million in seed funding, with contributions from notable investors like Matrix Partners and Y Combinator. This funding will be instrumental in hiring engineering and go-to-market staff, as well as covering computational costs for model training.
Lemon Slice-2's capabilities include live-streaming video at 20 frames per second and customization of avatar backgrounds and appearance. The company is also developing non-human characters and integrating voice generation technology, aiming to set a new standard in digital avatar realism and functionality.




