Publications

Abstract not provided.

Large scale non-intrusive inspection (NII) of commercial vehicles is being adopted in the U.S. at a pace and scale that will result in a commensurate growth in adjudication burdens at land ports of entry. The use of computer vision and machine learning models to augment human operator capabilities is critical in this sector to ensure the flow of commerce and to maintain efficient and reliable security operations. The development of models for this scale and speed requires novel approaches to object detection and novel adjudication pipelines. Here we propose a notional combination of existing object detection tools using a novel ensembling framework to demonstrate the potential for hierarchical and recursive operations. Further, we explore the combination of object detection with image similarity as an adjacent capability to provide post-hoc oversight to the detection framework. The experiments described herein, while notional and intended for illustrative purposes, demonstrate that the judicious combination of diverse algorithms can result in a resilient workflow for the NII environment.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

In this paper we apply Convolutional Neural Networks (CNNs) to the task of automatic threat detection, specifically conventional explosives, in security X-ray scans of passenger baggage. We present the first results of utilizing CNNs for explosives detection, and introduce a dataset, the Passenger Baggage Object Database (PBOD), which can be used by researchers to develop new threat detection algorithms. Using state-of-the-art CNN models and taking advantage of the properties of the Xray scanner, we achieve reliable detection of threats, with the best model achieving an AUC of the ROC of 0.95. We also explore heatmaps as a visualization of the location of the threat.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

This paper will investigate energy-efficiency for various real-world industrial computed-tomography reconstruction algorithms, both CPU- and GPU-based implementations. This work shows that the energy required for a given reconstruction is based on performance and problem size. There are many ways to describe performance and energy efficiency, thus this work will investigate multiple metrics including performance-per-watt, energy-delay product, and energy consumption. This work found that irregular GPU-based approaches1 realized tremendous savings in energy consumption when compared to CPU implementations while also significantly improving the performanceper- watt and energy-delay product metrics. Additional energy savings and other metric improvement was realized on the GPU-based reconstructions by improving storage I/O by implementing a parallel MIMD-like modularization of the compute and I/O tasks.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Abstract not provided.

To-date, the application of high-performance computing resources to Semantic Web data has largely focused on commodity hardware and distributed memory platforms. In this paper we make the case that more specialized hardware can offer superior scaling and close to an order of magnitude improvement in performance. In particular we examine the Cray XMT. Its key characteristics, a large, global shared-memory, and processors with a memory-latency tolerant design, offer an environment conducive to programming for the Semantic Web and have engendered results that far surpass current state of the art. We examine three fundamental pieces requisite for a fully functioning semantic database: dictionary encoding, RDFS inference, and query processing. We show scaling up to 512 processors (the largest configuration we had available), and the ability to process 20 billion triples completely in-memory.

We present an RDFS closure algorithm, specifically designed and implemented on the Cray XMT supercomputer, that obtains inference rates of 13 million inferences per second on the largest system configuration we used. The Cray XMT, with its large global memory (4TB for our experiments), permits the construction of a conceptually straightforward algorithm, fundamentally a series of operations on a shared hash table. Each thread is given a partition of triple data to process, a dedicated copy of the ontology to apply to the data, and a reference to the hash table into which it inserts inferred triples. The global nature of the hash table allows the algorithm to avoid a common obstacle for distributed memory machines: the creation of duplicate triples. On LUBM data sets ranging between 1.3 billion and 5.3 billion triples, we obtain nearly linear speedup except for two portions: file I/O, which can be ameliorated with the additional service nodes, and data structure initialization, which requires nearly constant time for runs involving 32 processors or more.