Trending...
- California: Governor Newsom, Superintendent Thurmond announce over $618 million to support another 458 community schools
- $4.3 Million Patent Application Waiver Fee Granted by FDA on New Drug Application Fee for Treatment Addressing Suicidal Depression & PTSD: NRX Pharma
- xREnergy up as much as +3,094,634% on first day listed on the XRP Ledger. Ticker : $XRE
LOS ANGELES - Californer -- Optical computing has been gaining wide interest for machine learning applications because of the massive parallelism and bandwidth of optics. Diffractive networks provide one such computing paradigm based on the transformation of the input light as it diffracts through a set of spatially-engineered surfaces, performing computation at the speed of light propagation without requiring any external power apart from the input light beam. Among numerous other applications, diffractive networks have been demonstrated to perform all-optical classification of input objects.
Researchers at the University of California, Los Angeles (UCLA), led by Professor Aydogan Ozcan, have introduced a 'time-lapse' scheme to significantly improve the image classification accuracy of diffractive optical networks on complex input objects. In this scheme, the object and/or the diffractive network are moved relative to each other during the exposure of the output detectors. Such a 'time-lapse' scheme has previously been used to achieve super-resolution imaging, for example, in security cameras, by capturing multiple images of a scene with lateral movements of the camera. Drawing inspiration from the success of time-lapse super-resolution imaging, UCLA researchers have utilized 'time-lapse diffractive networks' to achieve >62% blind testing accuracy on the all-optical classification of CIFAR-10 images, a publicly available dataset containing images of airplanes, cars, cats, etc. Their results achieved a significant improvement over the time-static diffractive optical networks.
More on The Californer
The same research group previously demonstrated ensemble learning of diffractive networks, where several diffractive networks operated in unison to improve image classification accuracy. However, with the incorporation of the 'time-lapse' scheme, it is possible to outdo an ensemble of >15 networks with a single standalone diffractive network, significantly reducing the footprint of the diffractive system while eliminating the complexities of physical alignment and synchronization of several individual networks. The researchers also explored the incorporation of ensemble learning into time-lapse image classification, which revealed >65% blind testing accuracy in classifying CIFAR-10 images.
For the physical implementation of the presented time-lapse classification scheme, the simplest method would exploit the natural jitter of the objects or the diffractive camera during imaging and allow the benefit of time-lapse to be reaped with no additional cost apart from a slight increase in the inference time due to detector signal integration during the jitter.
More on The Californer
This research on time-lapse image classification is a demonstration of utilizing the temporal degrees of freedom of optical fields for optical computing and presents a large step forward toward all-optical spatiotemporal information processing with compact, low-cost and passive materials.
Reference: Md Sadman Sakib Rahman and Aydogan Ozcan, Time-lapse image classification using a diffractive neural network, Advanced Intelligent Systems (2023). https://doi.org/10.1002/aisy.202200387
Researchers at the University of California, Los Angeles (UCLA), led by Professor Aydogan Ozcan, have introduced a 'time-lapse' scheme to significantly improve the image classification accuracy of diffractive optical networks on complex input objects. In this scheme, the object and/or the diffractive network are moved relative to each other during the exposure of the output detectors. Such a 'time-lapse' scheme has previously been used to achieve super-resolution imaging, for example, in security cameras, by capturing multiple images of a scene with lateral movements of the camera. Drawing inspiration from the success of time-lapse super-resolution imaging, UCLA researchers have utilized 'time-lapse diffractive networks' to achieve >62% blind testing accuracy on the all-optical classification of CIFAR-10 images, a publicly available dataset containing images of airplanes, cars, cats, etc. Their results achieved a significant improvement over the time-static diffractive optical networks.
More on The Californer
- Bennett Awards Designs Two Cohesive Custom Awards for 2025 Sci-Tech Academy Awards
- Poll Finds Overwhelming Opposition to Keeping Big Cats as Pets
- SM Telecom Expands AT&T Partnership to Deliver Cutting-Edge 5G+ Wireless Solutions New Collab Brings AT&T's Advanced 5G+ Technology to Cellphone
- Governor Newsom proclaims Older Californians Month
- Groundbreaking Launches 154-Acre Los Cerritos Wetlands Restoration in Long Beach – Single Largest Increase in Open Space in Long Beach in Decades
The same research group previously demonstrated ensemble learning of diffractive networks, where several diffractive networks operated in unison to improve image classification accuracy. However, with the incorporation of the 'time-lapse' scheme, it is possible to outdo an ensemble of >15 networks with a single standalone diffractive network, significantly reducing the footprint of the diffractive system while eliminating the complexities of physical alignment and synchronization of several individual networks. The researchers also explored the incorporation of ensemble learning into time-lapse image classification, which revealed >65% blind testing accuracy in classifying CIFAR-10 images.
For the physical implementation of the presented time-lapse classification scheme, the simplest method would exploit the natural jitter of the objects or the diffractive camera during imaging and allow the benefit of time-lapse to be reaped with no additional cost apart from a slight increase in the inference time due to detector signal integration during the jitter.
More on The Californer
- Don Barnhart Drops Unapologetically Hilarious Comedy Special "You Do You," On Open Bar Network
- Cygnet Theatre Announces The Cast and Creative Team of Rodgers and Hammerstein's Oklahoma!
- Shon Garage Door Repair Expands Trusted Services to San Diego, CA
- Nicky Dare Honors Indonesian Heritage at VAPI–Valley Asian and Pacific Islanders Cultural Festival
- California launches new AI-powered chatbot that provides wildfire resources in 70 languages
This research on time-lapse image classification is a demonstration of utilizing the temporal degrees of freedom of optical fields for optical computing and presents a large step forward toward all-optical spatiotemporal information processing with compact, low-cost and passive materials.
Reference: Md Sadman Sakib Rahman and Aydogan Ozcan, Time-lapse image classification using a diffractive neural network, Advanced Intelligent Systems (2023). https://doi.org/10.1002/aisy.202200387
Source: UCLA ITA
Filed Under: Science
0 Comments
Latest on The Californer
- Gravity to Bring 5-Minute EV Charging to 8 Sites Across Greater LA
- California: Governor Newsom issues statement on Pope Leo XIV, the first American Pope
- New poll shows high rates of employee burnout amid concerns over politics and personal finances
- Tessellations Appoints Luthern Williams as Head of School
- Aureli Construction Sets the Standard for Seamless Home Additions in Greater Boston
- Psychological Thriller "Killing Off Connor" To Open 34th IFS Film Fest After 12-years In Post
- Harvest Properties Acquires Two San Francisco Bay Area Self Storage Facilities for $44.2 Million
- California businesses in near-universal compliance with prohibition of intoxicating hemp products harmful to youth
- California: Governor Newsom announces upgrades to 21 state fish hatcheries to boost salmon populations
- Solaris Energy Infrastructure, Inc. (SEI) Investors Who Lost Money Have Opportunity to Lead Securities Fraud Lawsuit
- Risk Rater, Threat Assessment App, gives Users the Same Threat Evaluation as the Rich and Powerful
- Is it Really True That Tariffs Will Raise Car Insurance Rates?
- ScreenPoints Puts Film Investors in the Credits—and in the Money With New FinTech Platform
- Coastal Business Systems Wraps Up Successful 2025 Tech Show in Redding
- AdOcto Turns AirBnBs Into High-Impact Advertising Channels
- Zefr Announces Launch of Pre-Screen Brand Safety Solution for Google's Search Partner Network (SPN)
- Pathways to Adulthood Conference May 17 at Melville Marriott Honoring NYS Assembly Member Jodi Giglio, Suffolk County Legislator Nick Caracappa
- Adster Techologies awarded US Patent for breakthrough innovation in reducing latency in Ad Serving
- Flexi-View Lending Closes $5.05 Million Residential Acquisition Loan in Billings, Montana
- Robert Fabbio Inducted into the Austin Technology Council Hall of Fame