Jun Hwang

Nov 3, 2024 | nature.com | Kang Lee |Jun Hwang

AbstractMaterials scientists have taken a learn-from-nature approach to study the structure-property relationships of natural materials. Here we introduce a nature-inspired composite architecture showing a hierarchical assembly of granular-like building blocks with specific topological textures. The structural complexity of the resulting architecture is advanced by applying the concept of grain orientation internally to each building block to induce a tailored crack resistance.

Jul 27, 2023 | onlinelibrary.wiley.com | Hanna Lee |Jun Hwang |Seung Ho Song |Hyemi Han

Simulation of Convolutional-Neural-Network-Based Image Recognition The handwritten alphabet images were imported from the EMNIST dataset, which consisted of 124,800 training images and 20,800 test images. The intensities of test images, ranging from 0 to 255, were digitized to either 1 or 0 on the basis of whether the intensity was greater or less than the threshold value of 128, respectively. This process was used to prepare noiseless binary images for testing.