Diffractive networks light the way for optical image classification

January 13, 2021

Recently there has been a reemergence of interest in optical computing platforms for artificial intelligence-related applications. Optics/photonics is ideally suited for realizing neural network models because of the high speed, large bandwidth and high interconnectivity of optical information processing. Introduced by UCLA researchers, Diffractive Deep Neural Networks (D2NNs) constitute such an optical computing framework, comprising successive transmissive and/or reflective diffractive surfaces that can process input information through light-matter interaction. These surfaces are designed using standard deep learning techniques in a computer, which are then fabricated and assembled to build a physical optical network. Through experiments performed at terahertz wavelengths, the capability of D2NNs in classifying objects all-optically was demonstrated. In addition to object classification, the success of D2NNs in performing miscellaneous optical design and computation tasks, including e.g., spectral filtering, spectral information encoding, and optical pulse shaping have also been demonstrated.

In their latest paper published in Light: Science & Applications, UCLA team reports a leapfrog advance in D2NN-based image classification accuracy through ensemble learning. The key ingredient behind the success of their approach can be intuitively understood through the experiment of Sir Francis Galton (1822-1911), an English philosopher and statistician, who, while visiting a livestock fair, asked the participants to guess the weight of an ox. None of the hundreds of participants succeeded in guessing the weight. But to his astonishment, Galton found that the median of all the guesses came quite close - 1207 pounds, and was accurate within 1% of the true weight of 1198 pounds. This experiment reveals the power of combining many predictions in order to obtain a much more accurate prediction. Ensemble learning manifests this idea in machine learning, where an improved predictive performance is attained by combining multiple models.

In their scheme, UCLA researchers reported an ensemble formed by multiple D2NNs operating in parallel, each of which is individually trained and diversified by optically filtering their inputs using a variety of filters. 1252 D2NNs, uniquely designed in this manner, formed the initial pool of networks, which was then pruned using an iterative pruning algorithm, so that the resulting physical ensemble is not prohibitively large. The final prediction comes from a weighted average of the decisions from all the constituent D2NNs in an ensemble. The researchers evaluated the performance of the resulting D2NN ensembles on CIFAR-10 image dataset, which contains 60,000 natural images categorized in 10 classes and is an extensively used dataset for benchmarking various machine learning algorithms. Simulations of their designed ensemble systems revealed that diffractive optical networks can significantly benefit from the 'wisdom of the crowd'. For example, with an ensemble of 14 individually trained D2NNs, the researchers achieved 61.21% blind testing accuracy on CIFAR-10 dataset, which is ~16% higher than the average accuracy of the individual constituent D2NNs.

This research is led by Professor Aydogan Ozcan from the Electrical and Computer Engineering Department at UCLA, USA. This significantly improved inference and generalization performance achieved by D2NN ensembles marks a major advancement in closing the gap between optical neural networks and their digital counterparts. Together with the advances in the fabrication and assembly of nanoscale optical systems, the presented framework bears the promise for miniaturized, ultrafast machine learning solutions for a variety of applications, for example, all-optical object classification, diffraction-based optical computing hardware, and computational imaging tasks.

Light Publishing Center, Changchun Institute of Optics, Fine Mechanics And Physics, CAS

Related Weight Articles from Brightsurf:

How much postmenopause weight gain can be blamed on weight-promoting medications?
Abdominal weight gain, which is common during the postmenopause period, is associated with an array of health problems, including diabetes and heart disease.

Commercial weight management groups could support women to manage their weight after giving birth
Women who were overweight at the start of their pregnancy would welcome support after they have given birth in the form of commercial weight management groups, University of Warwick-led research has found.

Rollercoaster weight changes can repeat with second pregnancy, especially among normal-weight women
Everyone knows that gaining excess weight during one pregnancy is bad, but clinicians rarely consider weight gains and losses from one pregnancy to the next -- especially in normal-weight women.

Early and ongoing experiences of weight stigma linked to self-directed weight shaming
In a new study published today in Obesity Science and Practice, researchers at Penn Medicine and the University of Connecticut Rudd Center for Food Policy and Obesity surveyed more than 18,000 adults enrolled in the commercial weight management program WW International, and found that participants who internalized weight bias the most tended to be younger, female, have a higher body mass index (BMI), and have an earlier onset of their weight struggle

Being teased about weight linked to more weight gain among children, NIH study suggests
Youth who said they were teased or ridiculed about their weight increased their body mass by 33 percent more each year, compared to a similar group who had not been teased, according to researchers at the National Institutes of Health.

Association between weight before pregnancy, weight gain during pregnancy and adverse outcomes for mother, infant
An analysis that combined the results of 25 studies including nearly 197,000 women suggests prepregnancy body mass index (BMI) of the mother was more strongly associated with risk of adverse maternal and infant outcomes than the amount of gestational weight gain.

Study: Faster weight loss no better than slow weight loss for health benefits
Losing weight slowly or quickly won't tip the scale in your favor when it comes to overall health, according to new research.

What your choice of clothing says about your weight
It's commonly said that you can tell a great deal about a person by the clothes they wear.

Stand up -- it could help you lose weight
You might want to read this on your feet. A new study published today in the European Journal of Preventive Cardiology found that standing instead of sitting for six hours a day could prevent weight gain and help people to actually lose weight.

Cash for weight loss
A new study, published in the journal Social Science and Medicine, has shown that selling rewards programmes to participants entering a weight loss programme is a low cost strategy to increase both the magnitude and duration of weight loss.

Read More: Weight News and Weight Current Events
Brightsurf.com is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Amazon.com.