Three-dimensional shape sensing is critical for soft robotics, medical navigation, and intelligent perception systems, yet existing high-performance approaches typically rely on multicore fibers or multiple sensing channels. Researchers from Northwest University have developed a deformation-encoding strategy based on femtosecond-laser-inscribed misplaced orthogonal eccentric scatterers, enabling a standard single-mode fiber to reconstruct complex 3D shapes using only a single interrogation channel. The work establishes a new pathway toward miniaturized, high-density, and highly integrated optical fiber sensing systems.

Infrared adaptive optical systems that correct aberrations via deformable mirrors are critical for astronomy, defense, and industry fields. To this end, scientists in China invented a chain‑ring dual‑network chalcogenide glass that combine broadband infrared-transparent with polymer-like flexibility, which overcomes the long-standing material trade-off. Based on this novel material, they fabricated infrared deformable lenses with real‑time aberration correction, opening new pathways for infrared adaptive optical technology.

How can a spectrometer measure accurately on the scorching lunar farside? A new calibration framework for the Chang’e-6 Lunar Mineral Spectrometer corrected heat-induced spectral distortions, improving repeated measurement consistency by about 65%. The high-fidelity in situ data now bridge orbital observations, lunar surface measurements, and returned sample analyses, offering a clearer cross-scale view of the Chang’e-6 landing area.

Conventional second-harmonic generation microscopy is highly sensitive to symmetry, but it usually cannot directly read phase information. Researchers at Beijing Normal University developed an angular-polarization-engineered interferometric SHG microscope that uses a rotatable monolayer MoSe₂ reference placed between the objective and the sample. The system enables high-contrast phase-sensitive imaging of MoS₂ inversion twins and achieves an interference contrast as high as 0.96.

Motion correction in live imaging has long posed a critical challenge, as the rhythmic pulsations caused by breathing, heartbeat, and limb movements induce complex rigid and non-rigid deformations that severely compromise high-fidelity, long-term monitoring of large-scale neuronal populations. Now, through the development of an adaptive motion correction framework, a research team from Zhejiang University and Hangzhou City University has successfully overcome the accuracy and computational bottlenecks of traditional algorithms. The discovery is published in the journal PhotoniX.

A research team from the Beijing Institute of Technology, in collaboration with Minzu University of China, has proposed a paradigm-shifting hyperspectral imaging system that completely decouples spectral modulation from the detection process. Central to this system is a pixel-level spectrally programmable light-emitting diode that enables continuous, dynamic spectral tuning across the entire visible spectrum within a single, monolithic pixel, achieved via an innovative charge-generation layer architecture. Serving as an intelligent active illumination source, this programmable device works in concert with a standard commercial CMOS sensor and advanced reconstruction algorithms. The fully decoupled architecture achieves a remarkable 2 nm spectral resolution dictated entirely by the emitting device, alongside a 3072 × 2048 spatial resolution and a 60.9 fps raw frame rate—both dictated by the hardware limits of the CMOS sensor. By eliminating the need for light-dispersing or encoding optical elements, this system fundamentally shatters the traditional trade-offs among spatial, spectral, and temporal resolutions, avoids penalizing optical throughput, and circumvents complex micro-nano fabrication, thereby establishing a transformative path for high-performance hyperspectral imaging.

Dogs spontaneously follow the walking direction of a human figure represented by just a few moving dots, yet show no such directional bias when the dots depict another dog. In a study published in Cognitive Science Research, researchers found that dogs respond to directional cues from human biological motion as readily as they do to human head turns, suggesting that domestication may have tuned canine social perception to human movement.