Yining Hu

Ermyas Abebe, Yining Hu, Allison Irvin et al.

Permissioned ledger technologies have gained significant traction over the last few years. For practical reasons, their applications have focused on transforming narrowly scoped use-cases in isolation. This has led to a proliferation of niche, isolated networks that are quickly becoming data and value silos. To increase value across the broader ecosystem, these networks must seamlessly integrate with existing systems and interoperate with one another. A fundamental requirement for enabling crosschain communication is the ability to prove the validity of the internal state of a ledger to an external party. However, due to the closed nature of permissioned ledgers, their internal state is opaque to an external observer. This makes consuming and verifying states from these networks a non-trivial problem. This paper addresses this fundamental requirement for state sharing across permissioned ledgers. In particular, we address two key problems for external clients: (i) assurances on the validity of state in a permissioned ledger and (ii) the ability to reason about the currency of state. We assume an adversarial model where the members of the committee managing the permissioned ledger can be malicious in the absence of detectability and accountability. We present a formalization of the problem for state sharing and examine its security properties under different adversarial conditions. We propose the design of a protocol that uses a secure public ledger for providing guarantees on safety and the ability to reason about time, with at least one honest member in the committee. We then provide a formal security analysis of our design and a proof of concept implementation based on Hyperledger Fabric demonstrating the effectiveness of the proposed protocol.

Yining Hu, Suranga Seneviratne, Kanchana Thilakarathna et al.

Bitcoin is by far the most popular crypto-currency solution enabling peer-to-peer payments. Despite some studies highlighting the network does not provide full anonymity, it is still being heavily used for a wide variety of dubious financial activities such as money laundering, ponzi schemes, and ransom-ware payments. In this paper, we explore the landscape of potential money laundering activities occurring across the Bitcoin network. Using data collected over three years, we create transaction graphs and provide an in-depth analysis on various graph characteristics to differentiate money laundering transactions from regular transactions. We found that the main difference between laundering and regular transactions lies in their output values and neighbourhood information. Then, we propose and evaluate a set of classifiers based on four types of graph features: immediate neighbours, curated features, deepwalk embeddings, and node2vec embeddings to classify money laundering and regular transactions. Results show that the node2vec-based classifier outperforms other classifiers in binary classification reaching an average accuracy of 92.29% and an F1-measure of 0.93 and high robustness over a 2.5-year time span. Finally, we demonstrate how effective our classifiers are in discovering unknown laundering services. The classifier performance dropped compared to binary classification, however, the prediction can be improved with simple ensemble techniques for some services.

Tanner Bohn, Yining Hu, Charles X. Ling

We present an image preprocessing technique capable of improving the performance of few-shot classifiers on abstract visual reasoning tasks. Many visual reasoning tasks with abstract features are easy for humans to learn with few examples but very difficult for computer vision approaches with the same number of samples, despite the ability for deep learning models to learn abstract features. Same-different (SD) problems represent a type of visual reasoning task requiring knowledge of pattern repetition within individual images, and modern computer vision approaches have largely faltered on these classification problems, even when provided with vast amounts of training data. We propose a simple method for solving these problems based on the insight that removing peaks from the amplitude spectrum of an image is capable of emphasizing the unique parts of the image. When combined with several classifiers, our method performs well on the SD SVRT tasks with few-shot learning, improving upon the best comparable results on all tasks, with average absolute accuracy increases nearly 40% for some classifiers. In particular, we find that combining Relational Networks with this image preprocessing approach improves their performance from chance-level to over 90% accuracy on several SD tasks.

Tanner Bohn, Yining Hu, Jinhang Zhang et al.

We present a novel and effective technique for performing text coherence tasks while facilitating deeper insights into the data. Despite obtaining ever-increasing task performance, modern deep-learning approaches to NLP tasks often only provide users with the final network decision and no additional understanding of the data. In this work, we show that a new type of sentence embedding learned through self-supervision can be applied effectively to text coherence tasks while serving as a window through which deeper understanding of the data can be obtained. To produce these sentence embeddings, we train a recurrent neural network to take individual sentences and predict their location in a document in the form of a distribution over locations. We demonstrate that these embeddings, combined with simple visual heuristics, can be used to achieve performance competitive with state-of-the-art on multiple text coherence tasks, outperforming more complex and specialized approaches. Additionally, we demonstrate that these embeddings can provide insights useful to writers for improving writing quality and informing document structuring, and assisting readers in summarizing and locating information.

Yining Hu, Ahsan Manzoor, Parinya Ekparinya et al.

Banking as an essential service can be hard to access in remote, rural regions where the network connectivity is intermittent. Although micro-banking has been made possible by SMS or USSD messages in some places, their security flaws and session-based nature prevent them from a wider adoption. Global level cryptocurrencies enable low-cost, secure and pervasive money transferring among distributed peers, but are still limited in their ability to reach more people in remote communities. We proposed to take advantage of the delay-tolerant nature of blockchains to deliver banking services to remote communities that only connect to the broader Internet intermittently. Using a base station that offers connectivity within the local area, regular transaction processing is solely handled by blockchain miners. The bank only joins to process currency exchange requests, reward miners and track user balances when the connection is available. By distributing the verification and storage tasks among peers, our system design saves on the overall deployment and operational costs without sacrificing the reliability and trustwor- thiness. Through theoretical and empirical analysis, we provided insights to system design, tested its robustness against network disturbances, and demonstrated the feasibility of implementation on off-the-shelf computers and mobile devices.

Jagmohan Chauhan, Suranga Seneviratne, Yining Hu et al.

Recurrent neural networks (RNNs) have shown promising results in audio and speech processing applications due to their strong capabilities in modelling sequential data. In many applications, RNNs tend to outperform conventional models based on GMM/UBMs and i-vectors. Increasing popularity of IoT devices makes a strong case for implementing RNN based inferences for applications such as acoustics based authentication, voice commands, and edge analytics for smart homes. Nonetheless, the feasibility and performance of RNN based inferences on resources-constrained IoT devices remain largely unexplored. In this paper, we investigate the feasibility of using RNNs for an end-to-end authentication system based on breathing acoustics. We evaluate the performance of RNN models on three types of devices; smartphone, smartwatch, and Raspberry Pi and show that unlike CNN models, RNN models can be easily ported onto resource-constrained devices without a significant loss in accuracy.

Harini Kolamunna, Jagmohan Chauhan, Yining Hu et al.

The majority of available wearable devices require communication with Internet servers for data analysis and storage, and rely on a paired smartphone to enable secure communication. However, wearable devices are mostly equipped with WiFi network interfaces, enabling direct communication with the Internet. Secure communication protocols should then run on these wearables itself, yet it is not clear if they can be efficiently supported. In this paper, we show that wearable devices are ready for direct and secure Internet communication by means of experiments with both controlled and Internet servers. We observe that the overall energy consumption and communication delay can be reduced with direct Internet connection via WiFi from wearables compared to using smartphones as relays via Bluetooth. We also show that the additional HTTPS cost caused by TLS handshake and encryption is closely related to number of parallel connections, and has the same relative impact on wearables and smartphones.