"Smart TechSecure Voices"
About
SAVO (Secure Adversarial Voice Optimization)
generates adversarial data that disrupts AI training without affection regular users
SAVO inserts subtle modifications into voice data to proactively prevent unauthorized AI training. By generating data that preemptively blocks AI models from unauthorized learning, it ensures data integrity without compromising the original audio quality. Furthermore, its real-time prevention capability makes it adaptable for immediate application across various industries. With API and plugin support, SAVO provides scalability and ease of integration.
Security Solution Process
How Adversarial Data Disrupts AI Learning
Normal Dataset
(High Quality Data)
Generation of Contaminated Data
(Adversarial Samples)
Infiltration of Contaminated Data into the Normal Dataset
Distorted Outcomes in AI Model Training
Service
Planned Service Launch: January 2025
- 1. Real-time adversarial voice data generation service for voice data protection
- 2. High-quality voice data service designed to prevent AI training
- 3. Voice recognition and generation service based on adversarial voice data
- 4. High-quality voice data matching and trading service through identity verification
NEWS
Latest News
CONTACT US
Location and Contact
TEL : 02-6221-9120EMAIL : 912@9-1-2.comHead Office: Room 644, KU R&D Center, Korea University,
145 Anam-ro, Seongbuk-gu, Seoul, 02841, Republic of KoreaResearch Center: Room 303, MyWorkSpace, 597 Hoam-ro,
Gwanak-gu, Seoul, 08846, Republic of Korea
145 Anam-ro, Seongbuk-gu, Seoul, 02841, Republic of KoreaResearch Center: Room 303, MyWorkSpace, 597 Hoam-ro,
Gwanak-gu, Seoul, 08846, Republic of Korea