A team led by researchers Hao Wei and Xie Meilin at the Xi'an Institute of Optics and Precision Mechanics (XIOPM), Chinese Academy of Sciences, has proposed a resilient neural network-assisted model reference adaptive sliding mode control strategy to address deceptive attacks, actuator gain faults and coupled parametric uncertainties in multi-device systems. The results were published in IEEE Transactions on Industrial Electronics.
Researchers have proposed the SF-PULSE lensless imaging method, which can monitor living cells over long periods without staining, and successfully imaged and reconstructed glioblastoma cells.
Why we recommend it: The article demonstrates an imaging approach that replaces lenses with computational reconstruction, offering reference value for low-cost microscopy
Oct 8Thu
Thursday
Photonics.comOptics industryOptics and AIPicksAI score6161
Ghost imaging replaces pixelated sensors with correlation measurements between structured illumination and a single-pixel detector. Researchers combined an optical frequency comb with deep learning to achieve faster ghost imaging with higher fidelity.
Why we recommend it: The article shows the combination of frequency comb illumination and AI reconstruction, a typical case of optics and AI working together in computational imaging
Oct 5Mon
Monday
Optics & Photonics NewsOptics industryOptics and AIPicksAI score6363
The Ozcan laboratory at the University of California, Los Angeles developed a digital-optical hybrid system capable of identifying deepfake videos simultaneously across multiple channels.
Why we recommend it: The original article shows a new application of diffractive optical processors in AI vision tasks, a practical case of optical computing
Sep 30Wed
Wednesday
Shanghai Institute of Optics and Fine Mechanics (SIOM) · Research UpdatesOptics industryOptics and AIPicksAI score5959
The Joint Laboratory for High Power Laser and Physics at the Shanghai Institute of Optics and Fine Mechanics (SIOM) has proposed a laser spot analysis system based on an improved YOLOv8-seg to meet the need for rapid, precise alignment of optics along the full chain of high-power laser facilities, and has completed experimental validation.
Why we recommend it: The original paper demonstrates a deep learning segmentation model for laser facility alignment, a case of AI entering optical assembly and adjustment
Sep 16Wed
Wednesday
Light: Science & ApplicationsOptics industryOptics and AIPicksAI score6565
A paper in Light: Science & Applications combines dual combs with compressive ghost imaging and uses deep learning reconstruction to achieve hyperspectral imaging.
A team led by Guojun Wu at the Xi'an Institute of Optics and Mechanics has proposed PGF-ULB, an underwater optical beacon detection and localization framework combining physics guidance and geometric constraints, for AUV optical docking in complex underwater environments.
A team led by Peng Xie at the Shanghai Institute of Optics and Mechanics has made progress in multidimensional multiplexed optical computing, improving the parallelism of optical computing.
A team at the Optoelectronics Center of the Anhui Institute of Optics and Fine Mechanics has published two results in Remote Sensing, proposing the lightweight keypoint localization framework GeoAdapt and the frequency-domain low-rank adaptation segmentation method Freq-LoRA for aerial target sensing in ground-based optical remote sensing. GeoAdapt has only 5.95M parameters and achieves PCK@0.05D of 89.3%; Freq-LoRA requires only 559K trainable parameters and reaches a mean intersection-over-union of 0.904 across five simulated weather datasets, achieving 96.1% of full fine-tuning performance.
The Shanghai Institute of Optics and Fine Mechanics has published a review summarizing the applications, challenges and outlook of machine learning in inertial confinement fusion.
A team led by researcher Li Shaojuan at the Changchun Institute of Optics, Fine Mechanics and Physics has developed a vision-olfaction dual-modal neuromorphic device based on a MoS2/SnS2 floating-gate heterostructure. The device responds to both optical and gas signals and fuses them at the device level. The results were published in Advanced Materials. The device has an active area of 148 µm² and static power consumption of 3.4 µW. With the optical modality introduced, classification accuracy for 12 gases rose from 52.24% to 98.27%, while recognition accuracy for 51 gas mixtures reached 89.32%.
The Shanghai Institute of Optics and Mechanics has made progress in optically accelerated solving of large-spin-scale combinatorial optimization problems.
The Optical Remote Sensing Center of the Anhui Institute of Optics and Mechanics published a paper in IEEE TGRS proposing an attention-enhanced Kolmogorov-Arnold network (AKAN) method to retrieve aerosol optical depth from POSP polarization observations by the Gaofen-5B satellite. The team built more than 240,000 matched samples using AERONET ground-based data. The model achieved a coefficient of determination R² of 0.9336 and a root mean square error of 0.0492, with 94.25% of results falling within the expected error range. SHAP analysis showed that the sensitive bands and scattering angle features identified by the model are broadly consistent with radiative transfer principles.
The Optoelectronics Center of the Anhui Institute of Optics and Fine Mechanics published a paper in Analytical Chemistry describing CrisprDEM, a digital hydrogel fluorescence-enhanced microsphere detection system based on CRISPR/Cas12a that enables amplification-free relative quantification of nucleic acids. The system uses hydrogel microspheres to enrich fluorescent reporter molecules, boosting the fluorescence signal by a factor of about 1,000. A microfluidic chip then spreads the microspheres into a monolayer for imaging, and a K-means algorithm distinguishes positive from negative microspheres, with the proportion of positive microspheres used for quantification.
Teams led by Wang Binhao and Zhang Wenfu at the Xi'an Institute of Optics and Precision Mechanics have made a series of advances in silicon-based photonic interconnect chips.
The Xi'an Institute of Optics and Precision Mechanics and Northwest A&F University have combined AI with optical micromanipulation to achieve precise microparticle transport.