Cooper Union

About: Cooper Union is a education organization based out in New York, New York, United States. It is known for research contribution in the topics: Nonlinear system & Combustion. The organization has 414 authors who have published 509 publications receiving 11300 citations. The organization is also known as: The Cooper Union for the Advancement of Science and Art & Cooper Union College.

Challenges in Cardiac Tissue Engineering

Abstract: Cardiac tissue engineering aims to create functional tissue constructs that can reestablish the structure and function of injured myocardium. Engineered constructs can also serve as high-fidelity models for studies of cardiac development and disease. In a general case, the biological potential of the cell—the actual “tissue engineer”—is mobilized by providing highly controllable three-dimensional environments that can mediate cell differentiation and functional assembly. For cardiac regeneration, some of the key requirements that need to be met are the selection of a human cell source, establishment of cardiac tissue matrix, electromechanical cell coupling, robust and stable contractile function, and functional vascularization. We review here the potential and challenges of cardiac tissue engineering for developing therapies that could prevent or reverse heart failure.

Of Acceptable Risk: Science and the Determination of Safety

Abstract: One day, you will discover a new adventure and knowledge by spending more money. But when? Do you think that you need to obtain those all requirements when having much money? Why don't you try to get something simple at first? That's something that will lead you to know more about the world, adventure, some places, history, entertainment, and more? It is your own time to continue reading habit. One of the books you can enjoy now is of acceptable risk science and the determination of safety here.

EAD: Elastic-Net Attacks to Deep Neural Networks via Adversarial Examples

Abstract: Recent studies have highlighted the vulnerability of deep neural networks (DNNs) to adversarial examples — a visually indistinguishable adversarial image can easily be crafted to cause a well-trained model to misclassify. Existing methods for crafting adversarial examples are based on L 2 and L ∞ distortion metrics. However, despite the fact that L 1 distortion accounts for the total variation and encourages sparsity in the perturbation, little has been developed for crafting L 1 -based adversarial examples. In this paper, we formulate the process of attacking DNNs via adversarial examples as an elastic-net regularized optimization problem. Our elastic-net attacks to DNNs (EAD) feature L 1 -oriented adversarial examples and include the state-of-the-art L 2 attack as a special case. Experimental results on MNIST, CIFAR10 and ImageNet show that EAD can yield a distinct set of adversarial examples with small L 1 distortion and attains similar attack performance to the state-of-the-art methods in different attack scenarios. More importantly, EAD leads to improved attack transferability and complements adversarial training for DNNs, suggesting novel insights on leveraging L 1 distortion in adversarial machine learning and security implications of DNNs.

YOLO-LITE: A Real-Time Object Detection Algorithm Optimized for Non-GPU Computers

Abstract: This paper focuses on YOLO-LITE, a real-time object detection model developed to run on portable devices such as a laptop or cellphone lacking a Graphics Processing Unit (GPU). The model was first trained on the PASCAL VOC dataset then on the COCO dataset, achieving a mAP of 33.81% and 12.26% respectively. YOLO-LITE runs at about 21 FPS on a non-GPU computer and 10 FPS after implemented onto a website with only 7 layers and 482 million FLOPS. This speed is 3.8 × faster than the fastest state of art model, SSD MobilenetvI. Based on the original object detection algorithm YOLOV2, YOLO-LITE was designed to create a smaller, faster, and more efficient model increasing the accessibility of real-time object detection to a variety of devices.