Researchers unveil a passive all-optical device using neural networks for asymmetric image transmission and transformation.
Key Details
- 1Janus meta-imager uses cascaded metasurfaces and diffractive neural networks for image control.
- 2Device can pass images unchanged in one direction and transform them (e.g., letters to icons) in the reverse direction.
- 3Operates passively at high speed (~10 kHz) with minimal energy consumption.
- 4Device size is compact: 0.5 × 0.5 mm²; validated in the near-infrared (800 nm) regime.
- 5Potential applications include optical encryption and direction-dependent data storage.
Why It Matters
The Janus meta-imager exemplifies how deep learning and novel optics can enable new imaging functions without electronics or active power, suggesting future advances for secure medical image transport, coding, and all-optical processing architectures in radiology and related fields.
Source
EurekAlert
Related News
•EurekAlert
AI-Powered OCT Enables Rapid 'Optical Biopsy' for Early Endometrial Cancer Detection
A team at Washington University has developed a catheter-based 3D OCT system with AI to quickly and noninvasively detect early endometrial cancers.
•EurekAlert
AI Clinical Reasoning in Diagnostics and Digital Fatigue in Healthcare
Recent JMIR features explore large language models in clinical diagnostics and digital fatigue among healthcare professionals.
•EurekAlert
KAIST, MIT, Microsoft Develop Efficient AI Image Upsampling for Robotics
KAIST, MIT, and Microsoft have created 'Upsample Anything,' a training-free AI method to restore high-resolution visual data from compressed images with up to 16x improved GPU memory efficiency.