Company Performance Metrics
AIPARK's SaaS platform AiVATAR creates digital human videos that speak multiple languages by simply inputting text.
AiVATAR is a service that combines the following three technologies: - Text-To-Speech: AIPARK's proprietary artificial intelligence voice synthesis algorithm learns from datasets composed of text and audio and allows users
to output speech by only entering text, based on the trained AI voice synthesis algorithm. AIPARK has developed its own algorithms and possesses multiple TTS solutions. The AI voices implemented using this technology are being utilized in JTBC morning broadcasts, industrial broadcasts, and audiobook companies.
- Talking Head Generation: AIPARK's proprietary artificial intelligence video synthesis algorithm learns from tens of thousands of audio and facial data to output speaking videos regardless of the input voice. It has been trained on a large amount of data to naturally speak over 20 languages, including Korean, English, Japanese, and Chinese.
- Face Generation: This technology synthesizes and combines different faces to create completely new faces, generating virtual characters. It uses actual human faces rather than graphics for synthesis and re-creation, resulting in less visual discrepancy and freedom from portrait rights issues.
With the multilingual digital humans created by AiVATAR, users can experience unlimited applications such as news production, product promotion, and educational broadcasts.
