Chinese startup aims to surpass Sora with new AI video generator
By Ryan General
A Chinese startup is taking on OpenAI’s groundbreaking Sora with its own generative artificial intelligence (AI) model that creates videos from text input.
Key points:
Catch up:
- In February, Peking University and the Shenzhen-based AI company Rabbitpre also announced a similar text-to-video project called Open-Sora on Github.
The details:
- Shengshu-AI showcased Vidu on April 27 at the Zhongguancun Forum in Beijing. The demo shows Vidu being able to create highly detailed scenes with realistic physical properties like shadows and nuanced facial expressions.
- The company, founded in March 2023, claims it began developing its core technology before Sora’s release. Its chief scientist, Tsinghua University’s leading AI researcher Zhu Jun, reportedly published a related paper in September 2022. Inspired by Sora, the team accelerated Vidu’s development.
- The company is backed by investors like Ant Group and Baidu and has since raised over 100 million yuan ($13.8 million).
What’s next:
- Shengshu-AI said it aims to increase video duration and complexity, eventually hoping to match or surpass Sora’s current 60-second duration.
Share this Article
Share this Article
