Dynamic scene reconstruction is challenging because real-world scenes contain diverse types of motion, and no single motion representation consistently performs best across all scenarios. Researchers from Pusan National University developed a Mixture-of-Experts approaches that combine complementary dynamic representations to better capture diverse motions in dynamic 3D scenes. These approaches improve reconstruction quality while exploring different trade-offs between model flexibility, performance, and efficiency.

A research team found that scent compounds in household cleaning products — conventional and botanical essential oil-based — quickly react in the air, forming invisible nanoparticles that can travel deep into the lungs if inhaled. Several simple changes can reduce a person’s exposure while cleaning. The researchers will present their results at ACS Fall 2026.

Researchers from Zhejiang University and Zhejiang Advanced CNC Machine Tool Technology Innovation Center, China, have developed a Symmetry-Imbued Hamiltonian Neural Operator (SIHNO) for predicting stress fields in porous metamaterials. The proposed architecture combines rotation-equivariant geometric representations with a Hamiltonian-inspired energy structure to improve prediction accuracy and robustness under different structural orientations.

The framework integrates a Rotation-Steerable Lattice Projector (RSLP), a Hamiltonian-Fueled Propagator (HFP), and a slice-aware convolutional decoder to address two challenges in stress field prediction: orientation bias caused by fixed coordinate representations and non-physical stress leakage near void-solid interfaces.

In comprehensive numerical experiments based on finite-element-generated stress fields, SIHNO outperformed several established neural operator and physics-informed models across multiple evaluation metrics. The model also maintained millisecond-level inference speed, highlighting its potential for rapid stress prediction in additive manufacturing applications.

• KAIST and Yonsei University College of Medicine develop a miniaturized wireless brain implant capable of drug delivery, light stimulation, and IoT-based remote control

• A researcher in Chicago successfully operates a brain implant in Daejeon, Korea, in real time — demonstrating the feasibility of location-independent brain research

• Enables long-term, repeated experiments in freely moving animals, paving the way for advances in addiction, depression, and neurodegenerative disease research, as well as next-generation smart implantable therapies.