Skip To Content

Small Eyes & Smart Minds Incubator

Small Eyes & Smart Minds Incubator

4-6 October 2017
Washington, DC

Hosted by:
Rama Chellappa, University of Maryland, United States
Francisco Imai, Imaging Consultant, United States 
Ashok Veeraraghavan, Rice University, United States

Read more about this meeting: Day 1 and Day 2

View agenda here (pdf)
 

Background

Digital cameras are becoming omni-present and advancing technology has led to camera form factors that permit their use in numerous products, increasingly so in consumer products.  It is anticipated that they will continue to shrink, become more efficient, and offer novel new image processing features, on-board the camera. Reduced camera size will allow arrays of multiple cameras to be incorporated into a single device – a technological trend that is already emerging, particularly in surveillance and remote sensing applications.  Furthermore, individual cameras and multi-camera array devices will increasingly be networked across large, geographically separated areas.

Incubator Overview

OSA with the Society for Imaging Sciences and Technology,  brings this Incubator that will focus on the notion that camera is becoming a computer with eyes! In addition to exploring the fundamental limitations of optics as it relates to camera design (somewhat similar to the notion of Moore’s law in computing), the Incubator will explore traditional and novel applications in health, automotive industry, scientific imaging and virtual reality. The program will explore fundamental design issues in designing novel cameras from optics, materials, photonics and computational viewpoints. Is there something like the Moore’s law for designing cameras? What are the fundamental trade-offs in the emerging array of computational cameras?
 
We strongly believe that the next advances in this emerging area will arise out of collaboration between researchers working on metamaterials, photonics, flexible electronics, computational imaging, computer vision and graphics. One goal of the incubator will be to bring together researchers in these areas together in service of one application -- How do we make ultra-miniaturized, high-performance, smart cameras?

Scope and Featured Topics

This Incubator is designed to examine the expected evolution of digital imaging science and technology that is fueled by these trends in camera technology and to some extent driving it.  There are several areas where progress is expected which include the movement to smaller pixels which are on the order of a wavelength of light, on-chip digital processing, ways to effect curved focal planes, large angle & distortion free imagers, giga-pixel imagers, and stacked diode/multi-color focal planes, to mention a few.  Because of the prevalence of large numbers of cameras, networks of cameras will add new dimensions to our imaging capabilities, providing multi-views of objects and/or persistent situational awareness for both human, machine and internet needs.   Software processing is advancing to develop 3-D images, object tracking in arbitrary environments, structure from motion, multispectral views of scenes and data extraction. Lower cost electronics have stimulated computational and plenoptic sensing as well as feature/information identification.  These are just a few areas in processing which promise to significantly improve imaging capabilities in the future. These developments will give drive further advances in data science necessary to deal with the vast amount of image and video data that will be generated. This Incubator will:

  1. Attempt to extract the fundamental trade-offs between form-factor and image resolution/Field of view in the emerging area of tiny cameras.
  2. Explore the connection between the emerging technical areas of metamaterials, computational imaging and flexible electronics/sensors -- especially in the context of tiny cameras.
  3. Take an in-depth look as to what advances are in the pipeline are desired in both  camera sensor/hardware performance, form-factor, and other capabilities such as emerging processors and infrastructure bandwidth will subsequently afford which are not achievable today;
  4. Identify which of these will integrate well with innovations in algorithm development such as big data analytics and machine learning and new applications that have broad societal impact perhaps beyond the more familiar security, medical, entertainment and environmental applications;
  5. Explore applications in health sector, automotive sector, scientific imaging and virtual reality devices.

Meeting Sponsors

   

  
 

Image for keeping the session alive