Robust, long-range, passive depth perception and detection.
Resilient, long-range, passive detection and perception.
Light’s multi-view system captures an unprecedented level of information, redefining how automobiles see.
A missing piece in long-range depth perception
For vehicles to safely navigate the real world, they need to be able to perceive as humans do: a full picture with accurate depth throughout ranges. Lidar provides accurate information, but only up to a point, and with limited resolution. Radar detects when an object is in the far distance but it isn't sophisticated enough to discern whether its a truck or a barn. The range that radar is truly capable of is also often far less than claimed.
The hole that exists in long-range, accurate sensing for ADAS and AV is where Light comes in. We are developing an incredibly robust depth perception platform that will provide precise object detection, definition, and tracking through extended ranges. All in real-time.
Light's 3D perception system detects and perceives objects at ranges well beyond 100m, with the ability to detect relevant objects beyond half a kilometer. Built with a sophisticated system of precise camera controls, multi-view stereo, and computational software, our approach will deliver high-quality imaging and corresponding depth information in high-resolution.
Light's passive 3D depth-sensing uses multiple cameras to acquire data from the driving environment. Our algorithms accommodate various array and baseline configurations built to a manufacturer's specific requirements. While online calibration provides consistent results through the most rugged driving environments.
Small form factor.
Light's passive 3D depth-sensing utilizes multiple solid-state cameras to acquire data from the driving environment. This approach allows for power efficiency and elegant integration into a vehicle’s form factor. Our multi-view array calibration provides the ability to customize baselines to a manufacturer's specific requirements, and online calibration keeps the system working through any circumstance.
Light has proven expertise to solve the core challenges of passive-optics depth perception.
Multi-view, stereo, depth perception and detection
Multiple cameras create a detailed scene of distance based on the physics of multi-optic perception. Combined with computational algorithms, we can determine exact geometry and depth across the entire optical field of view.
Signal processing and machine learning
To fully understand the environment and identify and track objects within, we use a unique hybrid of machine learning and signal processing to determine depth in flat textureless regions and at object edges.
Powerful, high-resolution cameras. Small form-factor. Depth.
A high-resolution wide-angle video solution that fits into a small form factor and is energy efficient, which can also detect objects and people in 3D. Customizable to varying bandwidth needs and can be built using a variety of sensor configurations, including IR, and thermal.
Light's perception system allows robots to understand their operating environment deeply. Our depth-sensing system, configured with a sensor to meet your needs — panchromatic, thermal, IR, etc. is small, energy-efficient, and seamlessly integrates where space and power matter.
Light will completely revolutionize the world of mobile imaging.
Akshay Naheta, Partner
SoftBank Investment Advisors