High Driving Frequency (&gt;54 kHz) and Wide Scanning Angle (&gt;100 Degrees) MEMS Mirror Applying Secondary Resonance For 2K Resolution AR/MR Glasses

Mihara, Kensuke; Hanatani, Kaito; Ishida, T.; Komaki, Kazuki; Takayama, Ryoichi

doi:10.1109/mems51670.2022.9699653

Cited by 9 publications

(6 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Nevertheless, the majority of currently available SLMs exhibit sluggish performance, leading to significant latency [ 27 , 28 ], and the low refresh rate of SLMs significantly affects the throughput of BONN. Despite advances in MEMS technology, most SLMs still operate within a range of tens of kilohertz, significantly slower than electronic switches, which introduces substantial latency in BONN operations.…”

Section: Resultsmentioning

confidence: 99%

Bidirectional Optical Neural Networks Based on Free-Space Optics Using Lens Arrays and Spatial Light Modulator

2024

Micromachines

View full text Add to dashboard Cite

This paper introduces a novel architecture—bidirectional optical neural network (BONN)—for providing backward connections alongside forward connections in artificial neural networks (ANNs). BONN incorporates laser diodes and photodiodes and exploits the properties of Köhler illumination to establish optical channels for backward directions. Thus, it has bidirectional functionality that is crucial for algorithms such as the backpropagation algorithm. BONN has a scaling limit of 96 × 96 for input and output arrays, and a throughput of 8.5 × 1015 MAC/s. While BONN’s throughput may rise with additional layers for continuous input, limitations emerge in the backpropagation algorithm, as its throughput does not scale with layer count. The successful BONN-based implementation of the backpropagation algorithm requires the development of a fast spatial light modulator to accommodate frequent data flow changes. A two-mirror-like BONN and its cascaded extension are alternatives for multilayer emulation, and they help save hardware space and increase the parallel throughput for inference. An investigation into the application of the clustering technique to BONN revealed its potential to help overcome scaling limits and to provide full interconnections for backward directions between doubled input and output ports. BONN’s bidirectional nature holds promise for enhancing supervised learning in ANNs and increasing hardware compactness.

show abstract

Section: Resultsmentioning

confidence: 99%

Bidirectional Optical Neural Networks Based on Free-Space Optics Using Lens Arrays and Spatial Light Modulator

2024

Micromachines

View full text Add to dashboard Cite

show abstract

“…While we used an LCD as an example of an SLM in the proposed SOCNN, it is important to note that other types of SLMs can also be employed. However, SLMs based on micro-mirror or MEMS technology have certain drawbacks in system design [10][11][12][13][14]. They require input and output optical paths to be on the same side, which necessitates either sharing the optical path or using oblique light paths on the micro-mirror device.…”

Section: Methodsmentioning

confidence: 99%

“…In a Fourier-transform-based system, a laser and a spatial light modulator (SLM) are required to generate a coherent input. However, most of the currently available SLMs are slow and serially addressable, resulting in considerable latency [10][11][12][13][14]. This latency diminishes the advantages of massive parallelism in the optical neural network.…”

Section: Introductionmentioning

confidence: 99%

Scalable Optical Convolutional Neural Networks Based on Free-Space Optics Using Lens Arrays and a Spatial Light Modulator

2023

J. Imaging

View full text Add to dashboard Cite

A scalable optical convolutional neural network (SOCNN) based on free-space optics and Koehler illumination was proposed to address the limitations of the previous 4f correlator system. Unlike Abbe illumination, Koehler illumination provides more uniform illumination and reduces crosstalk. The SOCNN allows for scaling of the input array and the use of incoherent light sources. Hence, the problems associated with 4f correlator systems can be avoided. We analyzed the limitations in scaling the kernel size and parallel throughput and found that the SOCNN can offer a multilayer convolutional neural network with massive optical parallelism.

show abstract

“…Numerous studies have presented resonant 1D scanning micromirrors exhibiting large optical scan angles, sometimes higher than 100 • [1,2]. The in-depth characterization presented in this study, including frequency response measurements at different pressure levels and a quantification of dynamic deformation, enables a substantially more profound analysis and understanding of the microelectromechanical structures investigated.…”

Section: Introductionmentioning

confidence: 95%

A Comprehensive Characterization Procedure for Resonant MEMS Scanning Mirrors

Fleury,

Bainschab,

Mendicino

et al. 2024

Eurosensors 2023

View full text Add to dashboard Cite

show abstract

High Driving Frequency (>54 kHz) and Wide Scanning Angle (>100 Degrees) MEMS Mirror Applying Secondary Resonance For 2K Resolution AR/MR Glasses

Cited by 9 publications

References 11 publications

Bidirectional Optical Neural Networks Based on Free-Space Optics Using Lens Arrays and Spatial Light Modulator

Bidirectional Optical Neural Networks Based on Free-Space Optics Using Lens Arrays and Spatial Light Modulator

Scalable Optical Convolutional Neural Networks Based on Free-Space Optics Using Lens Arrays and a Spatial Light Modulator

A Comprehensive Characterization Procedure for Resonant MEMS Scanning Mirrors

Contact Info

Product

Resources

About