TruLens is a decentralized platform that enables trusted media verification using zero-knowledge proofs, IPFS, and Starknet. The platform allows users to capture images, attest them with metadata (timestamp, location), and verify their authenticity.
TruLens enables:
- Capture and verification of photos with cryptographic attestations
- Optional privacy-preserving GPS data verification
- Decentralized storage using IPFS/Pinata
- On-chain verification through Starknet
- Zero-knowledge proofs for private verification of capture conditions
┌─────────────┐ ┌─────────────┐ ┌─────────────┐
│ Frontend │────▶│ Backend │────▶│ IPFS │
│ (React/TS) │◀────│ (Node.js) │◀────│ (Pinata) │
└─────────────┘ └─────────────┘ └─────────────┘
│ │
│ │
▼ ▼
┌─────────────┐ ┌─────────────┐
│ Starknet │ │ Noir │
│ Contracts │ │ Circuits │
└─────────────┘ └─────────────┘
- Mobile-first design for in-the-field capture
- Image capture with metadata (GPS, timestamp)
- Wallet connection for Starknet interaction
- On-chain image attestation
- IPFS integration for decentralized storage
- Image sharing and saving capabilities
- IPFS storage via Pinata service
- Zero-knowledge proof generation and verification
- Persistent CID mapping for IPFS content
- RESTful API for frontend communication
- Zero-knowledge proofs using Noir language
- Flexible verification with or without GPS data
- Timestamp and location boundary verification
- Proof generation for privacy-preserving verification
- Smart contracts for on-chain attestation
- Decentralized verification and storage of attestation metadata
- Public verification of image authenticity
The platform addresses the challenge of verifying the authenticity of digital media, specifically:
- Proving an image was captured at a specific time
- Optionally verifying the capture location (GPS coordinates)
- Providing tamper-evident verification through cryptographic proofs
Our unique solution allows:
- Optional GPS verification (disabled GPS won't break verification)
- Privacy-preserving verification using zero-knowledge proofs
- Decentralized and persistent attestation records
The frontend is designed with a mobile-first approach, optimizing for:
- Field usage by journalists, field workers, and photographers
- Touch-friendly interface for quick capture
- Adaptive layout for various screen sizes
- Native device functionality (camera, GPS)
The Noir circuit allows verification without revealing sensitive data:
// Only verify GPS bounds if GPS is enabled
if gps_enabled_bool {
assert(latitude >= min_latitude_bound);
assert(latitude <= max_latitude_bound);
assert(longitude >= min_longitude_bound);
assert(longitude <= max_longitude_bound);
}All media and metadata are stored on IPFS through Pinata, with:
- Persistent gateway access
- Multiple fallback gateways
- Mapping between on-chain hashes and IPFS CIDs
- Node.js 16+
- PNPM package manager
- Noir v0.31.0 (for ZK circuit compilation)
- Starknet CLI/Account (for contract deployment)
# Clone the repository
git clone https://github.com/your-username/trulens.git
cd trulens
# Install dependencies
pnpm install
# Setup environment variables
cp backend/.env.example backend/.env
# Configure your Pinata API keys in .env# Start backend server
cd backend
pnpm start
# Start frontend development server
cd frontend
pnpm dev- Complete off-chain verification implementation
- Enhanced UI/UX for verification process
- Support for additional file formats (video, audio)
- Improved error handling and recovery
- Build native mobile app for better camera integration and provable image attestation
- World ID integration for sybil-resistant attestations and human verification
- Multi-chain support (Ethereum, Polygon)
- Integration with decentralized identity solutions
- Batch attestation support
- Enhanced metadata for specialized use cases
- Decentralized backend with P2P communication
- AI-resistant verification techniques
- Governance model for verification standards
- Industry-specific modules (journalism, insurance, legal)
-
Version Compatibility is critical in zero-knowledge circuits - matching versions of Noir, noir_js, and bb.js are required.
-
Witness Generation differs between Noir versions:
- v0.31.0 uses
nargo execute <witness_name> - Newer versions use
nargo prove
- v0.31.0 uses
-
Type Handling - Integer values must be properly converted to BigInt and GPS coordinates should be stored as integers (e.g., multiplied by 1,000,000).
-
Conditional Assertions in Noir provide more flexibility than hard assertions.
-
Mobile-First Development requires careful consideration of device capabilities and limitations.
Contributions are welcome! Please check out our contribution guidelines for details.
This project is licensed under the MIT License - see the LICENSE file for details.