Skip to main content

Dynamic Visual Cryptography

Think of it like a special movie that only human eyes can properly see, but with some unique security features.

Dynamic Visual Cryptography Image

Frame Generation​

Each "frame" of the movie is created by a one-time garbled circuit generated on the Interstellar blockchain.

Garbled circuits work like secure black boxes:

  • They receive encrypted random inputs
  • Process them without revealing internal logic
  • Output only the specific frames they’re designed to create

Every frame is cryptographically unique and cannot be reused.

Human Vision​

The system uses a human brain feature called persistence of vision:

  • Your eyes retain an image for ~1/30 of a second
  • When frames change rapidly (60–120 fps):
    • Your brain merges the sequence into one coherent image
    • Machines see static frames

This creates a channel that only humans can perceive and decode.

Security​

Intent data is embedded directly into the garbled circuit and never materializes in memory.

Compromising the system would require an attacker to chain together every one of the following:

  • Root-level compromise of the user's device
  • Advance knowledge of when the user will initiate an operation
  • Frame-perfect capture of every intent screen during the 2–4 second confirmation window
  • Pixel-exact reconstruction of the intent display
  • Faithful reproduction of native display timing (60–120 fps)
  • Real-time synchronization across all of the above

Even granting all of that, real-time probabilistic verification confirms that user responses correspond to the authentic, currently displayed sequence, and randomized per-frame visual watermarks defeat replay and precomputation. The economic ceiling on a successful attack is a single compromised transaction — the private key remains out of reach. The cost of mounting an attack of this sophistication vastly exceeds the payoff, making the ROI fundamentally negative: an adversary capable of clearing every bar above has far more lucrative targets elsewhere.

Analogy: The Flipbook​

Imagine a fast flipbook where:

  • Each page is generated by a secure black box
  • Each page looks like random noise
  • Only the human eye sees the real image when flipping quickly
  • Each book is a one-time use and randomized

Basic Simulation​

Below is a basic demonstration of visual cryptography.

When two same-sized images of apparently random black-and-white pixels are superimposed, the Wikipedia logo appears.

Visual Cryptographic Animation Demo

Interstellar's Dynamic Visual Cryptography​

This is the way Interstellar leverages human perssistence of vision with a basic visual cryptography scheme.

Superposed Frames Image

The garbled circuit manages the consecutive random frames display. This ensures that the circuit execution does not leak any secret information, such as the one-time code or the random keyboard topology associated with the transaction. The transaction message — amount, address, intent — is also embedded into the circuit, remaining inaccessible to attackers, preventing any modifications. This creates a secure visual channel between the blockchain and the user.

Garbled Circuits + Human Biology = Secure UI​

This secure display channel:

  • Prevents overlay attacks and clipper malware
  • Makes tampering visible
  • Works even on infected devices
  • Makes transaction spoofing ineffective

Instead of detecting malware, it creates an environment where spoofing is useless.

Deep Tech Advantage​

  • No need for kernel access or malware detection
  • Secure even on compromised devices
  • Simpler and clearer than traditional mobile EDR solutions

This technology enables hardware-wallet level security with nothing more than your smartphone.