The full agenda for the upcoming AutoSens event, to be launched in April, features over 30 virtual sessions including case studies, live demos and technical discussions.
Topics you can look forward to on the Agenda include:
- Tackling the limitations of camera-based perception
- RGB-IR image quality challenges and solutions
- Characteristics of image sensors for precise object detection in automotive ADAS systems
- Fast deep learning inference for autonomous driving and particle physics
- Building Automotive Solutions Using Edge AI/TinyML
- Industrializing High Performance Automotive LiDAR
- Environment and sensor digitalization for virtual driving in multiple level freeway interchange
- Solving the challenges of sensor model development in simulation
As well as leading-edge content from the finest brains in the industry, attendees can join the discussions from trailblazers on “Design for Manufacture challenges”, “The state of AI for ADAS” and “Extending the limits of CMOS image sensors.” These exciting sessions will feature speakers from Ford, OmniVision Technologies, Toyota and more.
Tickets on sale now for AutoSens
“RGB-IR image quality challenges and solutions” with Hongxin (Hans) Li, Principal ISP Architect, GEO Semiconductor Inc.
RGB-IR camera systems are capable of generating color and IR monochrome images simultaneously for Occupancy Monitoring System (OMS) and Driver Monitoring System (DMS) purposes. The ability to combine both color and IR video from a single camera reduces the cost and area required for these applications. As a new technology in automotive, this approach faces unique challenges, which need to be addressed from a system perspective, including areas related to optics, image sensors, and ISP algorithms.
Driver Monitoring Systems that are deployed today typically utilize a near IR sensor that works with active near IR illumination to provide consistent illumination of the driver in day and night. Since IR light shows much better contrast, at the iris, and pupil area, and can penetrate sunglasses, it is very suitable for blink and gaze detection. Recently occupant monitoring cameras are being added to interior camera applications, these cameras view the vehicle interior and need to provide color video. As a compact, more robust, and lower-cost alternative, the RGB-IR system combines the two cameras into one by using the RGB-IR sensor which contains both RGB and IR pixels, accompanied by additional signal processing to separate the RGB and IR images.
During the presentation at AutoSens, Hongxin (Hans) Li will guide the audience through a review of the architecture and design tradeoffs for a system of this nature and address some of the system challenges that are presented by this technology being applied to this application. System Challenges include Processing RGB-IR Video, Interference between multiple IR systems, Color and noise trade-off, HDR and Resolution.
Further innovation showcased
Additional trend-setting innovations to be highlighted during AutoSens this May, include “Active LiDAR” from AEye, which expands upon LiDAR’s intrinsic value to bring agility to the sensor, delivering ultra-long ranges, wide field of view, software configurability and instantaneous resolution.
Plus, Reality AI will discuss “How Edge AI + Sensors = Smarter Vehicles” and show some live examples of how to use Edge AI in different use cases, how to create Edge AI models, how to use AI to optimize sensor selection and placement, how to use AI to determine minimum component specifications, how to minimize the cost of data collection, and also how to generate sophisticated, explainable ML models based on sensor data, automatically.
Watch this space for the full Agenda to be announced next month.
Tickets on sale now for AutoSens