Dieses Stellenangebot ist nicht mehr verfügbar
Über
Proposals without relevant 3D avatar portfolio work will not be reviewed.
## Overview
We're building an AI companion mobile app (React Native) that features semi-realistic 3D avatars with real-time lip sync and facial expressions. We have a frontend developer building the app and a 3D artist creating the avatars — but we need an expert who can bridge the gap and ensure the avatars work properly in the app.
This is NOT a frontend development role. We need someone with deep expertise in , VRM avatars, and real-time facial animation who can guide our developer through implementation and troubleshoot issues.
**This engagement has two phases:**
## Phase 1: Get the First Two Avatars Working — $700 Fixed Price
You will work hands-on with our developer to get two VRM avatars fully functional in the app. This means rendering, lip sync, facial expressions, and performance — all working on mobile. This is the core deliverable.
**Scope:**
1. **VRM Avatar Rendering**
- Loading and displaying VRM avatars using @pixiv/three-vrm in React Native
- Performance optimization for mobile devices (older Android + iOS)
- Proper lighting, camera angles, and rendering quality
2. **Lip Sync Implementation**
- Mapping on-device TTS audio output to viseme blend shapes in real-time
- Low-latency viseme mapping to prevent the "dubbed movie" effect
- Smooth, natural-looking mouth movements synced with audio playback timing
3. **Facial Expressions**
- Triggering emotional expressions based on AI response sentiment
- Smooth transitions between expressions (happy, sad, thinking, etc.)
- Idle animations (blinking, subtle head movement)
- Proper state management within the React lifecycle to prevent memory leaks and crashes
4. **Performance & Optimization**
- Ensuring smooth 30+ FPS on mid-range mobile devices
- Texture and mesh optimization strategies
- Fallback handling for low-end devices
5. **Knowledge Transfer**
- Teaching our developer best practices for 3D in React Native so he can handle the remaining 4 avatars independently
- Code reviews on 3D-related implementations
- Documentation of the setup so it's repeatable
**Phase 1 Deliverables:**
- 2 VRM avatars rendering correctly in React Native
- Working lip sync tied to on-device TTS
- Facial expressions triggered by sentiment
- Idle animations (blink, subtle movement)
- 30+ FPS on mid-range devices
- Developer is trained and able to replicate the process for remaining avatars
## Phase 2: On-Call Consultation ($75/avatar if needed)
After Phase 1, our developer will handle the remaining 4 avatars independently using what he learned. If he runs into issues on any specific avatar, we'll bring you back at **$75 per avatar** to troubleshoot and resolve.
This is optional — if the knowledge transfer in Phase 1 is solid, we may not need Phase 2 at all.
## What We Have
- 6 semi-realistic VRM avatars (half-body) with full ARKit 52 blend shapes
- React Native mobile app (iOS + Android)
- On-device TTS for voice output (Kokoro/Piper)
- A frontend developer who needs guidance on 3D integration
## Required Skills
- Strong experience with and @pixiv/three-vrm
- Proven work with VRM or GLB avatar rendering in web or mobile
- Experience implementing lip sync (viseme-based) for 3D characters
- Understanding of ARKit blend shapes / Perfect Sync standard
- Familiarity with React Native or React context)
- Knowledge of mobile 3D performance optimization
- Experience managing state within the React lifecycle to prevent memory leaks
- Experience with Rhubarb Lip Sync or similar viseme generation tools
- Previous work on VTuber apps, avatar apps, or companion apps
- Experience with react-three-fiber or expo-three
## Nice to Have
- Understanding of on-device TTS systems (Kokoro, Piper)
## To Apply
Please include:
1. Portfolio showing or VRM avatar work (links to live demos preferred)
2. Specific examples of lip sync or facial animation you've implemented
3. Your experience with mobile 3D performance optimization
4. What is your preferred method for generating visemes from raw audio in a mobile environment?
Contract duration of less than 1 month.
Mandatory skills: 3D Modeling
Sprachkenntnisse
- English
Hinweis für Nutzer
Dieses Stellenangebot wurde von einem unserer Partner veröffentlicht. Sie können das Originalangebot einsehen hier.