Skip to main content

Implementation Roadmap

Phase 1: Proof of Concept​

Goal: Validate the core Proof of Face mechanism in isolation.

  • Implement hash-to-image pipeline (SHA-256 iterative expansion to 128x128 greyscale)
  • Integrate Haar Cascade face detection with scoring
  • Build a mining loop that finds face-producing nonces
  • Benchmark: how many nonces per second can be tested?
  • Benchmark: what is the distribution of face scores across random inputs?
  • Calibrate initial difficulty threshold based on benchmarks
  • Determinism test suite: verify identical results across platforms
  • Simple proof verification: given a challenge + miner + nonce, verify the face score

Deliverable: A standalone mining CLI that finds faces in hashes and reports scores.

Phase 2: Rollup with Mining​

Goal: Deploy the Proof of Face mechanism as a Cosmos SDK module on a Rollkit sovereign rollup.

  • Scaffold Cosmos SDK chain with Rollkit
  • Implement x/facecoin module:
    • MsgSubmitProofOfFace transaction type
    • On-chain face detection verification (gocv or pure-Go Haar Cascade)
    • Challenge rotation logic
    • Difficulty adjustment algorithm
    • FACE token minting on valid proof
    • NFT minting on valid proof
  • MsgTransferNFT transaction type
  • Genesis state configuration
  • Connect to Celestia testnet (Mocha)
  • Mining client that watches chain state and submits proofs
  • Basic block explorer / face gallery web UI

Deliverable: A testnet rollup on Celestia where participants can mine FACE, submit proofs, and view the face gallery.

Phase 3: Polish and Testnet​

Goal: Production-quality testnet with full feature set.

  • Determinism test suite across platforms (ensure all nodes compute identical scores)
  • Mining client optimizations (parallel nonce search, GPU-accelerated detection)
  • RPC endpoints for querying face images, NFT metadata, mining stats
  • Wallet integration (Keplr or custom)
  • Face gallery web app with colorized rendering
  • CLIP-based aesthetic scoring (optional, off-chain) for NFT quality ranking
  • Load testing and performance tuning
  • Public testnet launch

Deliverable: Public testnet with mining client, wallet, and gallery.

Phase 4: Bridge and Ecosystem​

Goal: Connect Facecoin to the broader crypto ecosystem.

  • Deploy wFACE (ERC-20) contract on Ethereum
  • Deploy FacecoinNFT (ERC-721) contract on Ethereum
  • Implement bridge (Axelar, Lazybridging, or custom lock-and-mint)
  • Bridge UI for transferring assets between chains
  • Deploy SPL token on Solana (wFACE)
  • Deploy Metaplex NFT on Solana
  • Wallet integration guides (MetaMask, Phantom)

Deliverable: Working bridges to Ethereum and Solana with UI.

Phase 5: Mainnet​

Goal: Launch the production network.

  • Security audit of application module and bridge contracts
  • Migrate from Celestia testnet to mainnet
  • Performance testing under load
  • Mainnet launch
  • Bridge activation
  • Documentation and community onboarding

Deliverable: Live mainnet with active mining and bridging.

Open Research Questions​

  1. Optimal image size: Is 128x128 the right tradeoff between detection quality and generation speed? Benchmarking will inform this.

  2. Detector choice finalization: Should the protocol support multiple detectors (with the specific detector ID recorded in the proof), or commit to exactly one?

  3. Color mapping for NFTs: What aesthetic approach produces the most visually interesting colorized faces? This is subjective and could benefit from community input.

  4. Sequencer decentralization: When and how to transition from a centralized sequencer to shared/decentralized sequencing (Astria, Espresso, etc.).

  5. Governance: How should protocol parameters (difficulty adjustment period, proof rate target, detector parameters) be changed over time? On-chain governance? Hard forks?

  6. CLIP integration: Should a CLIP-based aesthetic score be added as optional NFT metadata? As a chain parameter? This could enable "proof of beautiful face" at higher difficulty tiers.

  7. Mobile mining: Face detection on mobile devices is well-supported (MediaPipe was designed for it). Could Facecoin support a mobile mining client for broader participation?