Philomuse develops 3D interactive communication technology that integrates real-time facial expressions, posture, eye-contact, and spatial sound to simulate co-presence. This platform enables users to engage with highly intelligent, emotionally responsive AI chatbots that understand and remember conversational context. The core offering delivers rich, non-verbal language and emotional cues within immersive digital interactions.
Funding
Funding not disclosed
Founders
Product
Problem
Current metaverse and virtual communication platforms lack the nuance of in-person interactions, failing to capture non-verbal cues like facial expressions, posture, and spatial audio, which are crucial for authentic social connection. This absence limits user engagement and the sense of "presence" in virtual environments.
Solution
Presence is developing an AI and AR platform that leverages real-time facial expression recognition, posture tracking, and spatial audio to create more immersive and engaging social interactions. Their technology aims to bridge the gap between virtual and physical communication by accurately translating non-verbal cues into the metaverse. By reconstructing 3D environments and avatars that respond realistically to user behavior, Presence seeks to foster a stronger sense of co-presence and connection among users. The platform supports AI roleplay, enabling users to interact with super-intelligent chatbots that understand and remember them.
Target Audience
The primary target audience includes individuals and organizations seeking more engaging and realistic social experiences in the metaverse, such as virtual event organizers, remote collaboration teams, and users of social VR platforms.
Features
- Real-time facial expression recognition for avatar animation
- Posture tracking to reflect body language in the virtual environment
- Spatial audio processing to simulate realistic soundscapes
- 3D reconstruction using particle projectors and depth sensors
- AI roleplay capabilities with intelligent chatbots
- Mobile inference frameworks requiring less than 1G FLOPS