Core Mission & Overview

The Photonics Machine Intelligence Lab bridges the worlds of advanced optics and modern artificial intelligence. The lab develops next-generation hardware platforms that use light to process data, while simultaneously applying machine learning tools to solve complex photonic problems and control opto-biological systems.

Research Domains

• Computational Optics: Designing optical platforms that act as hardware accelerators for machine learning tasks, using linear and non-linear optical processes to perform machine learning tasks.
• Machine Learning for Photonics: Employing AI and machine learning algorithms to control light fields and optimize optical data processing for applications in imaging and healthcare.
• Opto-Biological Neural Interfaces: Investigating the use of targeted optical fields to control and interface with biological neural networks, driving the development of brain-machine interfaces and opto-biological computers.
• Advanced Beam Optics: Creating theoretical and experimental methods to construct exotic, tailored optical beams for precision microscopy, optical tweezing, and computing.

Capabilities

• All-Optical Inference Engines: Developing high-speed, wave-based hardware platforms capable of executing machine learning tasks directly in the optical domain.
• Nonlinear Diffractive Processing: Utilizing nonlinear optical wave mixing and diffraction to process complex data distributions with high efficiency.
• Structured Light Synthesis: Designing and controlling structured light beams with tailored parameters to manipulate spatial modes and interact with physical or biological systems.
• High-Throughput Optical Computing: Building custom multi-mode optical fiber setups and crystal configurations optimized for advanced vector-matrix operations.

Industry Collaboration & Opportunities

• Next-Generation AI Hardware: Partnering with tech companies to develop all-optical accelerators for fast, low-power machine learning inference and neural network layers.
• Advanced Healthcare & Medical Imaging: Collaborating on intelligent optical data processing systems to improve resolution and speed in clinical diagnostics.
• Potential Market Segments: AI Hardware Acceleration, Optical Computing, Telecommunications, Biotechnology & Neuroengineering, and Advanced Medical Imaging.

Selected Publications

• All-Optical Inference Engine Based on a Shape-Optimized Multimode Optical Fiber (Physical Review Applied) – Demonstrating how tailored fiber geometries can perform complex data inference completely in the optical domain.
• All-Optical Inference Using Structured Light Beams (Physical Review Applied) – Showcasing how spatial light profiles can be leveraged to compute machine learning tasks without electronic overhead.
• Optical Computing Using Nonlinear Optical Diffraction (Machine Learning in Photonics) – Outlining framework solutions for executing high-fidelity matrix operations via non-linear diffraction.

Contact & Collaboration

For industrial partnerships or collaborative research inquiries, please contact Ramot or Prof. Alon Bahabad directly at the Department of Physical Electronics, Tel Aviv University.