Tsung-Yi Ho

Bio: Tsung-Yi Ho is an academic researcher from National Tsing Hua University. The author has contributed to research in topics: Biochip & Routing (electronic design automation). The author has an hindex of 34, co-authored 288 publications receiving 3790 citations. Previous affiliations of Tsung-Yi Ho include Indian Statistical Institute & National Taiwan University.

Error Recovery in Cyberphysical Digital Microfluidic Biochips

Abstract: Droplet-based digital microfluidics technology has now come of age, and software-controlled biochips for healthcare applications are starting to emerge. However, today's digital microfluidic biochips suffer from the drawback that there is no feedback to the control software from the underlying hardware platform. Due to the lack of precision inherent in biochemical experiments, errors are likely during droplet manipulation; error recovery based on the repetition of experiments leads to wastage of expensive reagents and hard-to-prepare samples. By exploiting recent advances in the integration of optical detectors (sensors) into a digital microfluidics biochip, we present a physical-aware system reconfiguration technique that uses sensor data at intermediate checkpoints to dynamically reconfigure the biochip. A cyberphysical resynthesis technique is used to recompute electrode-actuation sequences, thereby deriving new schedules, module placement, and droplet routing pathways, with minimum impact on the time-to-response.

CloudLeak: Large-Scale Deep Learning Models Stealing Through Adversarial Examples.

Abstract: Cloud-based Machine Learning as a Service (MLaaS) is gradually gaining acceptance as a reliable solution to various real-life scenarios. These services typically utilize Deep Neural Networks (DNNs) to perform classification and detection tasks and are accessed through Application Programming Interfaces (APIs). Unfortunately, it is possible for an adversary to steal models from cloud-based platforms, even with black-box constraints, by repeatedly querying the public prediction API with malicious inputs. In this paper, we introduce an effective and efficient black-box attack methodology that extracts largescale DNN models from cloud-based platforms with near-perfect performance. In comparison to existing attack methods, we significantly reduce the number of queries required to steal the target model by incorporating several novel algorithms, including active learning, transfer learning, and adversarial attacks. During our experimental evaluations, we validate our proposed model for conducting theft attacks on various commercialized MLaaS platforms hosted by Microsoft, Face++, IBM, Google and Clarifai. Our results demonstrate that the proposed method can easily reveal/steal large-scale DNN models from these cloud platforms. The proposed attack method can also be used to accurately evaluates the robustness of DNN based MLaaS classifiers against theft attacks.

A fast routability- and performance-driven droplet routing algorithm for digital microfluidic biochips

Abstract: As the microfluidic technology advances, the design complexity of digital microfluidic biochips (DMFB) are expected to explode in the near future. One of the most critical challenges for DMFB design is the droplet routing problem, which schedules the movement of each droplet in a time-multiplexed manner. In this paper, we propose a fast routability- and performance-driven droplet router for DMFBs. The main contributions of our work are: (1) a global moving vector analysis for constructing preferred routing tracks to minimize the number of used unit cells; (2) an entropy-based equation to determine the routing order of droplets for better routability; (3) a routing compaction technique by dynamic programming to minimize the latest arrival time of droplets. Experimental results show that our algorithm achieves 100% routing completion for all test cases on three Benchmark Suites while the previous algorithms are not. In addition to routability, compared with the state-of-the-art high-performance routing on the Benchmark Suite I [3], the experimental results still show that our algorithm performed better in runtime by 40%, reduced the latest arrival time by 21%, reduced the used unit cells by 10%. Furthermore, experiment results on Benchmark Suite II and III are also very promising. Based on the evaluation of three Benchmark Suites, our algorithm demonstrates the efficiency and robustness of handling complex droplet routing problem over the existing algorithms.

Robust Adversarial Objects against Deep Learning Models

Abstract: Previous work has shown that Deep Neural Networks (DNNs), including those currently in use in many fields, are extremely vulnerable to maliciously crafted inputs, known as adversarial examples. Despite extensive and thorough research of adversarial examples in many areas, adversarial 3D data, such as point clouds, remain comparatively unexplored. The study of adversarial 3D data is crucial considering its impact in real-life, high-stakes scenarios including autonomous driving. In this paper, we propose a novel adversarial attack against PointNet++, a deep neural network that performs classification and segmentation tasks using features learned directly from raw 3D points. In comparison to existing works, our attack generates not only adversarial point clouds, but also robust adversarial objects that in turn generate adversarial point clouds when sampled both in simulation and after construction in real world. We also demonstrate that our objects can bypass existing defense mechanisms designed especially against adversarial 3D data.

A Reagent-Saving Mixing Algorithm for Preparing Multiple-Target Biochemical Samples Using Digital Microfluidics

Abstract: Recent advances in digital microfluidics have led to the promise of miniaturized laboratories, with the associated advantages of high sensitivity and less human-induced errors. Front-end operations such as sample preparation play a pivotal role in biochemical laboratories, and in applications in biomedical engineering and life science. For fast and high-throughput biochemical applications, preparing samples of multiple target concentrations sequentially is inefficient and time-consuming. Therefore, it is critical to concurrently prepare samples of multiple target concentrations. In addition, since reagents used in biochemical reactions are expensive, reagent-saving has become an important consideration in sample preparation. Prior work in this area does not address the problem of reagent-saving and concurrent sample preparation for multiple target concentrations. In this paper, we propose the first reagent-saving mixing algorithm for biochemical samples of multiple target concentrations. The proposed algorithm not only minimizes the consumption of reagents, but it also reduces the number of waste droplets and the sample preparation time by preparing the target concentrations concurrently. The proposed algorithm is evaluated on both real biochemical experiments and synthetic test cases to demonstrate its effectiveness and efficiency. Compared to prior work, the proposed algorithm can achieve up to 41% reduction in the number of reagent droplets and waste droplets, and up to 50% reduction in sample preparation time.

Higher-order organization of complex networks

Abstract: Networks are a fundamental tool for understanding and modeling complex systems in physics, biology, neuroscience, engineering, and social science. Many networks are known to exhibit rich, lower-order connectivity patterns that can be captured at the level of individual nodes and edges. However, higher-order organization of complex networks—at the level of small network subgraphs—remains largely unknown. Here, we develop a generalized framework for clustering networks on the basis of higher-order connectivity patterns. This framework provides mathematical guarantees on the optimality of obtained clusters and scales to networks with billions of edges. The framework reveals higher-order organization in a number of networks, including information propagation units in neuronal networks and hub structure in transportation networks. Results show that networks exhibit rich higher-order organizational structures that are exposed by clustering based on higher-order connectivity patterns.

Design Of Analog Cmos Integrated Circuits

Abstract: Thank you very much for downloading design of analog cmos integrated circuits. Maybe you have knowledge that, people have look hundreds times for their favorite novels like this design of analog cmos integrated circuits, but end up in malicious downloads. Rather than reading a good book with a cup of coffee in the afternoon, instead they cope with some malicious virus inside their laptop. design of analog cmos integrated circuits is available in our book collection an online access to it is set as public so you can download it instantly. Our digital library saves in multiple countries, allowing you to get the most less latency time to download any of our books like this one. Merely said, the design of analog cmos integrated circuits is universally compatible with any devices to read.

A Survey of Neuromorphic Computing and Neural Networks in Hardware.

Abstract: Neuromorphic computing has come to refer to a variety of brain-inspired computers, devices, and models that contrast the pervasive von Neumann computer architecture This biologically inspired approach has created highly connected synthetic neurons and synapses that can be used to model neuroscience theories as well as solve challenging machine learning problems The promise of the technology is to create a brain-like ability to learn and adapt, but the technical challenges are significant, starting with an accurate neuroscience model of how the brain works, to finding materials and engineering breakthroughs to build devices to support these models, to creating a programming framework so the systems can learn, to creating applications with brain-like capabilities In this work, we provide a comprehensive survey of the research and motivations for neuromorphic computing over its history We begin with a 35-year review of the motivations and drivers of neuromorphic computing, then look at the major research areas of the field, which we define as neuro-inspired models, algorithms and learning approaches, hardware and devices, supporting systems, and finally applications We conclude with a broad discussion on the major research topics that need to be addressed in the coming years to see the promise of neuromorphic computing fulfilled The goals of this work are to provide an exhaustive review of the research conducted in neuromorphic computing since the inception of the term, and to motivate further work by illuminating gaps in the field where new research is needed