Peter Robinson

Pradipta Biswas, Pilar Orero, Manohar Swaminathan et al.

Augmented, virtual and mixed reality technologies offer new ways of interacting with digital media. However, such technologies are not well explored for people with different ranges of abilities beyond a few specific navigation and gaming applications. While new standardization activities are investigating accessibility issues with existing AR/VR systems, commercial systems are still confined to specialized hardware and software limiting their widespread adoption among people with disabilities as well as seniors. This proposal takes a novel approach by exploring the application of user model-based personalization for AR/VR systems to improve accessibility. The workshop will be organized by experienced researchers in the field of human computer interaction, robotics control, assistive technology, and AR/VR systems, and will consist of peer reviewed papers and hands-on demonstrations. Keynote speeches and demonstrations will cover latest accessibility research at Microsoft, Google, Verizon and leading universities.

Mauricio Soto-Gomez, Peter Robinson, Carlos Cano et al.

Many real-world problems are naturally modeled as heterogeneous graphs, where nodes and edges represent multiple types of entities and relations. Existing learning models for heterogeneous graph representation usually depend on the computation of specific and user-defined heterogeneous paths, or in the application of large and often not scalable deep neural network architectures. We propose Het-node2vec, an extension of the node2vec algorithm, designed for embedding heterogeneous graphs. Het-node2vec addresses the challenge of capturing the topological and structural characteristics of graphs and the semantic information underlying the different types of nodes and edges of heterogeneous graphs, by introducing a simple stochastic node and edge type switching strategy in second order random walk processes. The proposed approach also introduces an ''attention mechanism'' to focus the random walks on specific node and edge types, thus allowing more accurate embeddings and more focused predictions on specific node and edge types of interest. Empirical results on benchmark datasets show that Hetnode2vec achieves comparable or superior performance with respect to state-of-the-art methods for heterogeneous graphs in node label and edge prediction tasks.

Peter Robinson, Raghavendra Ramesh

The General Purpose Atomic Crosschain Transaction protocol allows composable programming across multiple Ethereum blockchains. It allows for inter-contract and inter-blockchain function calls that are both synchronous and atomic: if one part fails, the whole call execution tree of function calls is rolled back. The protocol operates on existing Ethereum blockchains without modification. It works for both public permissioned and consortium blockchains. Additionally, the protocol is expected to work across heterogeneous blockchains other than Ethereum. This paper describes the protocol, analyses it in terms of Gas usage and Finalised Block Periods for three scenarios: reading a value from one blockchain to another, writing a value from one blockchain to another, and a trade finance system involving five contracts on five blockchains with a complex call execution tree, and provides an initial security analysis that shows that the protocol has Safety and Liveness properties.

Peter Robinson

Atomic Crosschain Transaction technology allows composable programming across permissioned Ethereum blockchains. It allows for inter-contract and inter-blockchain function calls that are both synchronous and atomic: if one part fails, the whole call graph of function calls is rolled back. This paper analyses the processing overhead of using this technique compared to using multiple standard non-atomic single blockchain transactions. The additional processing is analysed for three scenarios involving multiple blockchains: the Hotel - Train problem, Supply Chain with Provenance, and an Oracle. The technology is shown to reduce the performance of Hyperledger Besu from 375 tps to 39.5 tps if all transactions are instigated on one node, or approaching 65.2 tps if the transactions are instigated on a variety of nodes, for the Hotel-Train scenario.

Peter Robinson, Raghavendra Ramesh

The Layer 2 Atomic Cross-Blockchain Function Calls protocol allows composable programming across Ethereum blockchains. It allows for inter-contract and inter-blockchain function calls that are both synchronous and atomic: if one part fails, the whole call graph of function calls is rolled back. Existing atomic cross-blockchain function call protocols are Blockchain Layer 1 protocols, which require changes to the blockchain platform software to operate. Blockchain Layer 2 technologies such as the one described in this paper require no such changes. They operate on top of the infrastructure provided by the blockchain platform software. This paper introduces the protocol and a more scalable variant, provides an initial safety and liveness analysis, and presents the expected overhead of using this technology when compared to using multiple non-atomic single blockchain transactions. The overhead is analysed for three scenarios involving multiple blockchains: the Hotel and Train problem, Supply Chain with Provenance, and an Oracle. The protocol is shown to provide 93.8 or 186 cross-blockchain function calls per second for the Hotel and Train scenario when there are many travel agencies, for the standard and scalable variant of the protocol respectively, given the Ethereum client, Hyperledger Besu's performance of 375 tps, assuming a block period of one second, and assuming all transactions take the same amount of time to execute as the benchmark transactions.

Peter Robinson

Crosschain communications allows information to be communicated between blockchains. Consensus in the context of crosschain communications relates to how participants on one blockchain are convinced of the state of a remote blockchain. It describes how parties associated with a source blockchain come to agreement and communicate with a destination blockchain such that information from the source blockchain can be trusted. This paper surveys crosschain communications protocols, presenting them based on the top-level usage scenarios they are trying to meet: value swapping, crosschain messaging, and blockchain pinning. It analyses how each protocol achieves crosschain consensus, what trust assumptions are made, their ability to operate successfully in Permissionless and Permissioned blockchains contexts, and whether the protocol delivers atomic updates across blockchains.

Peter Robinson, Raghavendra Ramesh, John Brainard et al.

Atomic Crosschain Transaction technology allows composable programming across private Ethereum blockchains. It allows for inter-contract and inter-blockchain function calls that are both synchronous and atomic: if one part fails, the whole call graph of function calls is rolled back. It is not based on existing techniques such as Hash Time Locked Contracts, relay chains, block header transfer, or trusted intermediaries. BLS Threshold Signatures are used to prove to validators on one blockchain that information came from another blockchain and that a majority of the validators of that blockchain agree on the information. Coordination Contracts are used to manage the state of a Crosschain Transaction and as a repository of Blockchain Public Keys. Dynamic code analysis and signed nested transactions are used together with live argument checking to ensure execution only occurs if the execution results in valid state changes. Contract Locking and Lockability enable atomic updates.

J. Luc Peterson, Ben Bay, Joe Koning et al.

With the growing complexity of computational and experimental facilities, many scientific researchers are turning to machine learning (ML) techniques to analyze large scale ensemble data. With complexities such as multi-component workflows, heterogeneous machine architectures, parallel file systems, and batch scheduling, care must be taken to facilitate this analysis in a high performance computing (HPC) environment. In this paper, we present Merlin, a workflow framework to enable large ML-friendly ensembles of scientific HPC simulations. By augmenting traditional HPC with distributed compute technologies, Merlin aims to lower the barrier for scientific subject matter experts to incorporate ML into their analysis. In addition to its design, we describe some example applications that Merlin has enabled on leadership-class HPC resources, such as the ML-augmented optimization of nuclear fusion experiments and the calibration of infectious disease models to study the progression of and possible mitigation strategies for COVID-19.

Peter Robinson

Atomic Crosschain Transaction technology allows composable programming across private Ethereum blockchains. It allows for inter-contract and inter-blockchain function calls that are both synchronous and atomic: if one part fails, the whole call graph of function calls is rolled back. Traditional Ethereum contract functions can limit which accounts can call them by specialised application program logic. This is important as it allows application developers to specify which callers can execute functions that update contract state. In this paper we introduce the strategy required to restrict which contracts on one blockchain can call a function in a contract that is deployed on another blockchain. We show that validating the Originating Blockchain Id (the blockchain the crosschain function call started on), From Blockchain Id, and From Account provides contracts with certainty that a function call came from a specific contract on a specific blockchain.

Sam Ade Jacobs, Brian Van Essen, David Hysom et al.

Training deep neural networks on large scientific data is a challenging task that requires enormous compute power, especially if no pre-trained models exist to initialize the process. We present a novel tournament method to train traditional as well as generative adversarial networks built on LBANN, a scalable deep learning framework optimized for HPC systems. LBANN combines multiple levels of parallelism and exploits some of the worlds largest supercomputers. We demonstrate our framework by creating a complex predictive model based on multi-variate data from high-energy-density physics containing hundreds of millions of images and hundreds of millions of scalar values derived from tens of millions of simulations of inertial confinement fusion. Our approach combines an HPC workflow and extends LBANN with optimized data ingestion and the new tournament-style training algorithm to produce a scalable neural network architecture using a CORAL-class supercomputer. Experimental results show that 64 trainers (1024 GPUs) achieve a speedup of 70.2 over a single trainer (16 GPUs) baseline, and an effective 109% parallel efficiency.

Peter Robinson

A Coordination Blockchain is a blockchain with the task of coordinating activities of multiple private blockchains. This paper discusses the pros and cons of using Ethereum MainNet, the public Ethereum blockchain, as a Coordination Blockchain. The requirements Ethereum MainNet needs to fulfil to perform this role are discussed within the context of Ethereum Private Sidechains, a private blockchain technology which allows many blockchains to be operated in parallel, and allows atomic crosschain transactions to execute across blockchains. Ethereum MainNet is a permissionless network which aims to offer strong authenticity, integrity, and non-repudiation properties, that incentivises good behaviour using crypto economics. This paper demonstrates that Ethereum MainNet does deliver these properties. It then provides a comprehensive review of the features of Ethereum Private Sidechains, with a focus on the potential usage of Coordination Blockchains for these features. Finally, the merits of using Ethereum MainNet as a Coordination Blockchain are assessed. For Ethereum Private Sidechains, we found that Ethereum MainNet is best suited to storing long term static data that needs to be widely available, such as the Ethereum Registration Authority information. However, due to Ethereum MainNet's probabilistic finality, it is not well suited to information that needs to be available and acted upon immediately, such as the Sidechain Public Keys and Atomic Crosschain Transaction state information that need to be accessible prior to the first atomic crosschain transaction being issued on a sidechain. Although this paper examined the use of Ethereum MainNet as a Coordination Blockchain within reference to Ethereum Private Sidechains, the discussions and observations of the typical tasks a Coordination blockchain may be expected to perform are applicable more widely to any multi-blockchain system.

Sandra Johnson, Peter Robinson, Kishore Atreya et al.

Supply chains lend themselves to blockchain technology, but certain challenges remain, especially around invoice financing. For example, the further a supplier is removed from the final consumer product, the more difficult it is to get their invoices financed. Moreover, for competitive reasons, retailers and manufacturers do not want to disclose their supply chains. However, upstream suppliers need to prove that they are part of a `stable' supply chain to get their invoices financed, which presents the upstream suppliers with huge, and often unsurmountable, obstacles to get the necessary finance to fulfil the next order, or to expand their business. Using a fictitious supply chain use case, which is based on a real world use case, we demonstrate how these challenges have the potential to be solved by combining more advanced and specialised blockchain technologies with other technologies such as Artificial Intelligence. We describe how atomic crosschain functionality can be utilised across private blockchains to retrieve the information required for an invoice financier to make informed decisions under uncertainty, and consider the effect this decision has on the overall stability of the supply chain.

Peter Robinson, David Hyland-Wood, Roberto Saltini et al.

Public blockchains such as Ethereum and Bitcoin do not give enterprises the privacy they need for many of their business processes. Consequently consortiums are exploring private blockchains to keep their membership and transactions private. Ethereum Private Sidechains is a private blockchain technology which allows many blockchains to be operated in parallel. Communication is needed between Ethereum Private Sidechains to allow a function in a contract on one sidechain to execute function calls which return values from, or update the state of, another sidechain. We propose a crosschain technique which allows transactions to be executed atomically across sidechains, introduce a new mechanism for proving values across sidechains, describe a transaction locking mechanism which works in the context of blockchain to enable atomic transactions, and a methodology for providing a global time-out across sidechains. We outline the programming model to be used with this technology and provide as an example, a variable amount atomic swap contract for exchanging value between sidechains. Although this paper presents Atomic Crosschain Transaction technology in the context of Ethereum Private Sidechains, we discuss how this technology can be readily applied to many blockchain systems to provide cross-blockchain transactions.

Sandra Johnson, Peter Robinson, John Brainard

There appears to be an insatiable desire for spawning new bespoke blockchains to harness the functionality provided by blockchain technologies, resulting in a constant stream of blockchain start-up companies entering the market with their own unique vision and mission. Some target a particular niche market such as supply chain and financial services, while others strive to differentiate themselves from the increasingly saturated market by offering new functionality. This dynamic and constantly changing blockchain ecosystem makes it very challenging to keep abreast of all the latest breakthroughs and research. It is evident that there is also a growing desire to collaborate with others developing blockchain solutions, which brings new impetus to blockchain interoperability research. We review the strategies that some key players in the blockchain ecosystem have implemented, or are proposing to develop, to satisfy this increasing demand for cross-chain communication and transactions between sidechains. Interoperability presents a complex and challenging stumbling block to the wider uptake of blockchain technology. We find that although there is a plethora of blockchains and interoperability implementations, or proposals, at a higher level of abstraction there is only a handful of approaches. However, the way they are implemented can differ quite substantially. We present a summary of the reviews we conducted in a table for ease of comparing and contrasting.

Peter Robinson, John Brainard

Public blockchains such as Ethereum and Bitcoin provide transparency and accountability, and have strong non-repudiation properties, but fall far short of enterprise privacy requirements for business processes. Consequently consortiums are exploring private blockchains to keep their membership and transactions private. However, private blockchains do not provide adequate protection against potential collusion by consortium members to revert the state of the blockchain. To countenance this, the private blockchain state may be "pinned" to a tamper resistant public blockchain. Existing solutions offering pinning to the public blockchain would reveal the transaction rate of the private blockchain, and do not provide a mechanism to contest the validity of a pin. Moreover, they require that all transactions and members of the private blockchain be revealed. These challenges are hampering the wider adoption of private blockchain technology. We describe the primary author's `Anonymous State Pinning approach', which overcomes these limitations and present a security proof to demonstrate pins can be challenged without compromising these properties. We perform a gas cost analysis of the implementation to estimate the operating cost of this technology, which shows that pinning a private blockchain at the rate of one pin per hour would cost US$508 per year. A hierarchical pinning approach is proposed which would allow many private blockchains to pin to a management blockchain which would then pin to Ethereum MainNet. This approach saves money, but at the cost of increased finality times.

Peter Robinson

Decentralised random number generation algorithms suffer from the Last Actor Problem, in which the last participant to reveal their share can manipulate the generated random value by withholding their share. This paper proposes an encrypted share threshold scheme which prevents this attack.

Peter Robinson

The Enterprise Ethereum Client Specification by the Enterprise Ethereum Alliance defines the requirements which Ethereum Clients offering private smart contract capabilities should comply with. This specification though ground breaking, misses some important blockchain requirements and does not fully consider the requirements of Ethereum Clients offering Private Sidechain capabilities. This paper presents the case for Private Sidechains and defines requirements to be complied with to deliver this technology. The capabilities of three blockchain clients have been analysed based on the requirements: Quorum, Parity, and Hyperledger Fabric. Quorum and Hyperledger Fabric operate as private consortium blockchains where as Parity delivers private transaction capabilities on top of Ethereum MainNet. These differing approaches has led to different strengths and weaknesses which has resulted in each client not complying with one or more key requirement. In particular, none of the reviewed blockchain clients support the ability to determine bootstrap information to establish on-demand blockchains and none of the clients support secure management and pinning from Ethereum MainNet. This paper presents Ethereum Private Sidechains and a range of technologies which allow it to deliver on complex sidechain requirements. Ethereum Registration Authorities are presented, which allow entities which have not previously interacted to securely obtain information to bootstrap a sidechain, and a Management and Pinning strategy is described which allows the state of a sidechain to be securely pinned to Ethereum MainNet without compromising privacy.

Peter Robinson

Message digest algorithms are one of the underlying building blocks of blockchain platforms such as Ethereum. This paper analyses situations in which the message digest collision resistance property can be exploited by attackers. Two mitigations for possible attacks are described: longer message digest sizes make attacks more difficult; and, including timeliness properties limits the amount of time an attacker has to determine a hash collision.

Erroll Wood, Tadas Baltrusaitis, Louis-Philippe Morency et al.

We present GazeDirector, a new approach for eye gaze redirection that uses model-fitting. Our method first tracks the eyes by fitting a multi-part eye region model to video frames using analysis-by-synthesis, thereby recovering eye region shape, texture, pose, and gaze simultaneously. It then redirects gaze by 1) warping the eyelids from the original image using a model-derived flow field, and 2) rendering and compositing synthesized 3D eyeballs onto the output image in a photorealistic manner. GazeDirector allows us to change where people are looking without person-specific training data, and with full articulation, i.e. we can precisely specify new gaze directions in 3D. Quantitatively, we evaluate both model-fitting and gaze synthesis, with experiments for gaze estimation and redirection on the Columbia gaze dataset. Qualitatively, we compare GazeDirector against recent work on gaze redirection, showing better results especially for large redirection angles. Finally, we demonstrate gaze redirection on YouTube videos by introducing new 3D gaze targets and by manipulating visual behavior.

Heng Yang, Xuhui Jia, Chen Change Loy et al.

The problem of face alignment has been intensively studied in the past years. A large number of novel methods have been proposed and reported very good performance on benchmark dataset such as 300W. However, the differences in the experimental setting and evaluation metric, missing details in the description of the methods make it hard to reproduce the results reported and evaluate the relative merits. For instance, most recent face alignment methods are built on top of face detection but from different face detectors. In this paper, we carry out a rigorous evaluation of these methods by making the following contributions: 1) we proposes a new evaluation metric for face alignment on a set of images, i.e., area under error distribution curve within a threshold, AUC$_α$, given the fact that the traditional evaluation measure (mean error) is very sensitive to big alignment error. 2) we extend the 300W database with more practical face detections to make fair comparison possible. 3) we carry out face alignment sensitivity analysis w.r.t. face detection, on both synthetic and real data, using both off-the-shelf and re-retrained models. 4) we study factors that are particularly important to achieve good performance and provide suggestions for practical applications. Most of the conclusions drawn from our comparative analysis cannot be inferred from the original publications.

Heng Yang, Renqiao Zhang, Peter Robinson

In this paper we present a method for localisation of facial landmarks on human and sheep. We introduce a new feature extraction scheme called triplet-interpolated feature used at each iteration of the cascaded shape regression framework. It is able to extract features from similar semantic location given an estimated shape, even when head pose variations are large and the facial landmarks are very sparsely distributed. Furthermore, we study the impact of training data imbalance on model performance and propose a training sample augmentation scheme that produces more initialisations for training samples from the minority. More specifically, the augmentation number for a training sample is made to be negatively correlated to the value of the fitted probability density function at the sample's position. We evaluate the proposed scheme on both human and sheep facial landmarks localisation. On the benchmark 300w human face dataset, we demonstrate the benefits of our proposed methods and show very competitive performance when comparing to other methods. On a newly created sheep face dataset, we get very good performance despite the fact that we only have a limited number of training samples and a set of sparse landmarks are annotated.

Heng Yang, Wenxuan Mou, Yichi Zhang et al.

In this paper we propose a supervised initialization scheme for cascaded face alignment based on explicit head pose estimation. We first investigate the failure cases of most state of the art face alignment approaches and observe that these failures often share one common global property, i.e. the head pose variation is usually large. Inspired by this, we propose a deep convolutional network model for reliable and accurate head pose estimation. Instead of using a mean face shape, or randomly selected shapes for cascaded face alignment initialisation, we propose two schemes for generating initialisation: the first one relies on projecting a mean 3D face shape (represented by 3D facial landmarks) onto 2D image under the estimated head pose; the second one searches nearest neighbour shapes from the training set according to head pose distance. By doing so, the initialisation gets closer to the actual shape, which enhances the possibility of convergence and in turn improves the face alignment performance. We demonstrate the proposed method on the benchmark 300W dataset and show very competitive performance in both head pose estimation and face alignment.

Erroll Wood, Tadas Baltrusaitis, Xucong Zhang et al.

Images of the eye are key in several computer vision problems, such as shape registration and gaze estimation. Recent large-scale supervised methods for these problems require time-consuming data collection and manual annotation, which can be unreliable. We propose synthesizing perfectly labelled photo-realistic training data in a fraction of the time. We used computer graphics techniques to build a collection of dynamic eye-region models from head scan geometry. These were randomly posed to synthesize close-up eye images for a wide range of head poses, gaze directions, and illumination conditions. We used our model's controllability to verify the importance of realistic illumination and shape variations in eye-region training data. Finally, we demonstrate the benefits of our synthesized training data (SynthesEyes) by out-performing state-of-the-art methods for eye-shape registration as well as cross-dataset appearance-based gaze estimation in the wild.

Björn Schuller, Erik Marchi, Simon Baron-Cohen et al.

Individuals with Autism Spectrum Conditions (ASC) have marked difficulties using verbal and non-verbal communication for social interaction. The running ASC-Inclusion project aims to help children with ASC by allowing them to learn how emotions can be expressed and recognised via playing games in a virtual world. The platform includes analysis of users' gestures, facial, and vocal expressions using standard microphone and web-cam or a depth sensor, training through games, text communication with peers, animation, video and audio clips. We present the state of play in realising such a serious game platform and provide results for the different modalities.