CandyVoice

CandyVoice develops a modular technology portfolio for real-time voice processing :

• Voice bio-metric data anonymization

• Audio deepfake detection

• Voice imitation

• Advanced noise filtering

• Lost-frame recovery (PLC)

Our proprietary algorithms are ultra-low-latency, lightweight, and designed for constrained environments - from tactical radio-communications to edge devices.

Real-time voice imitation for immersive worlds:
Our real-time voice imitation technology enables users to embody their avatar with a personalized vocal identity in real time, with natural expressiveness preserved. In virtual worlds, XR environments, and multiplayer gaming, voice is the most intimate vector of presence and social interaction. CandyVoice brings user-centric immersion where the voice seamlessly matches the visual identity, enhancing embodiment, emotional engagement, and cross-platform interoperability. This opens new possibilities for multisensory experiences, accessible communication, and truly persistent digital identities across interconnected virtual spaces.

Relevance to Horizon Europe priorities:

1. Trustworthy AI & data sovereignty (DATA-06): Voice anonymization enabling GDPR-compliant synthetic voice data generation; deepfake detection for AI content authenticity.

2. Immersive experiences & Web 4.0 (HUMAN-01): Voice skins, avatar vocal embodiment, and interoperable voice identities for virtual worlds and XR.

3. Telco-Edge-Cloud infrastructure (DATA-08): Ultra-low-latency noise filtering and packet-loss concealment for resilient voice communications across distributed edge environments; voice quality enhancement for mission-critical applications in mobility, healthcare, and industrial IoT.

4. Applied AI challenges (DIGITAL-EMERGING-04): Voice interfaces for healthcare, automotive, and industrial use cases

We are seeking: Consortium coordinators and system integrators looking for a specialized voice technology partner with mature, deployment-ready solutions (TRL 6-8).