Unleashing AI on Light
Traditional cameras capture just one slice of the 7D plenoptic function. BlurLens captures it all—turning every photon into multi-dimensional data through our N-sensor advantage. While others fight blur, we decode it: each blur disc encodes object distance, and our multi-sensor array reads these optical signatures like a 3D barcode, unlocking significantly more informationthan single-sensor systems, with high-precision depth capture across a broad operating range.
Learn MoreLight as a Layered Signal
BlurLens treats every photon as a multi-dimensional messenger. While traditional cameras flatten light into 2D, we decode the full 7D plenoptic function. Each blur disc encodes precise depth information—our multi-sensor array reads these optical signatures like a 3D barcode:
- Pixel-Level Highly accurate depth from defocus patterns, material properties from polarimetric analysis
- Object-Level Motion vectors from temporal defocus shifts
- Scene-Level Super-resolution through Nyquist-compliant sensor overlap
- System-Level Very high dynamic range without motion artifacts
Our physics-informed neural architecture unlocks a far greater information capacity than single-sensor systems. By embracing blur as information rather than noise, we unlock capabilities impossible with traditional optics.
Technical Breakthroughs
- • Proven signal-to-noise improvements in low-light conditions
- • Unified loss functions balancing depth, motion, and material accuracy
- • Real-time processing on standard mobile NPUs
- • Post-capture refocus from f/1.0 to f/22 equivalent
Protected by patents pending. Confidentiality required.
Rewriting the Laws of Computational Optics
We’re hiring physicists who dream in PSFs, ML engineers fluent in physics-informed CNNs, and hardware hackers who see calibration drift as a personal insult. If you’ve ever argued with a ray-tracing simulation at 2AM or wondered why cameras still can’t see like humans do—let’s fix it.
Have questions about open roles or our work culture? Drop us a line at [email protected].
The Missing Link in Computational Imaging
While others layer AI onto legacy camera designs, BlurLens reimagines imaging from first principles. Our patent-pending hardware and AI co-design creates a new category between traditional optics and pure computational methods. It is dramatically more affordable than LIDAR for autonomous vehicles and substantially faster than laser scanners for industrial inspection, while delivering notably better low-light performance.
vs. Light Field Cameras
No resolution tradeoffs
Super-resolution imaging gains
Ultra-wide dynamic range
vs. LiDAR
Passive operation
Works in fog/rain/snow
Significant cost reduction
vs. Traditional AI
Greater information capacity
Physics-guided models
Real-time on NPUs
With multiple patents filed around multi-sensor defocus analysis and AI-native optics, we're positioned to enable:
- Industrial metrology
- Cinematic refocusing
- AR/VR depth sensing
- Smartphone imaging
We're on track to license our technology across industries while maintaining a clear path to consumer-scale deployment at a rapid pace.