Memory-Augmented Generative Adversarial Networks for Anomaly Detection.

We propose a memory-augmented deep learning model for semisupervised anomaly detection (AD). While many traditional AD methods focus on modeling the distribution of normal data, additional constraints in the modeling process are needed to distinguish between normal and abnormal data. The proposed model, named memory augmented generative adversarial networks (MEMGAN), is coupled with external memory units through attentional operations.

Multi-eigenmode control for high material contrast in bimodal and higher harmonic atomic force microscopy.

High speed imaging and mapping of nanomechanical properties in atomic force microscopy (AFM) allows the observation and characterization of dynamic sample processes. Recent developments involve several cantilever frequencies in a multifrequency approach. One method actuates the first eigenmode for topography imaging and records the excited higher harmonics to map nanomechanical properties of the sample.

Aggrecan nanoscale solid-fluid interactions are a primary determinant of cartilage dynamic mechanical properties.

Poroelastic interactions between interstitial fluid and the extracellular matrix of connective tissues are critical to biological and pathophysiological functions involving solute transport, energy dissipation, self-stiffening and lubrication. However, the molecular origins of poroelasticity at the nanoscale are largely unknown. Here, the broad-spectrum dynamic nanomechanical behavior of cartilage aggrecan monolayer is revealed for the first time, including the equilibrium and instantaneous moduli and the peak in the phase angle of the complex modulus.