For annotation in cancer genomic medicine, oncologists have to refer to various knowledge bases worldwide and retrieve all information (e.g., drugs, clinical trials, and academic papers) related to a gene variant. However, oncologists find it difficult to search these knowledge bases comprehensively because there are multiple paraphrases containing abbreviations and foreign languages in their terminologies including diseases, drugs, and genes. In this paper, we propose a novel search method considering deep paraphrases, which helps oncologists retrieve essential annotation resources swiftly and effortlessly. Our method recursively finds paraphrases based on paraphrase corpora, expands a source document, and finally generates a paraphrase lattice. The proposed method also feedbacks beneficial information regarding the paraphrases applied for a search, which is useful for selecting search results and considering a query for the succeeding search. The results of an experiment demonstrated that our method could retrieve important annotation information that could not be retrieved using a conventional search system and simple paraphrasing. Additionally, annotation experts evaluated our method and found it to be practical.

Worldwide, many surveillance systems are in operation for crime deterrence purposes. An effective system should be characterized by requiring low-power consumption, a small storage capacity, and little human effort. Multi-stream tracking on field programmable gate array (FPGA) is important for such surveillance systems. In this paper, we propose multi-stream tracking hardware that can extract moving objects and their motion vectors from a multi-stream received from 64 cameras in real time. The key technology for multi-stream processing is as follows. (1) In order to avoid maintaining the background, we apply a frame difference method. Moreover, the flows of object are calculated by block matching. The flows are effective for analyzing human motion. (2) In order to avoid a bus bottleneck and memory contention in the communication between processing elements (PEs), synchronous shift data transfer (SSDT), which transfers data in the same direction for all PEs, is applied. In this paper, an extended SSDT is proposed for communication between PEs when multi-blocks are processed in one PE. (3) C++ based integrated control code development tool is shown. Control code written in C++ language can easily be assembled and verified by the tool. We implemented the proposed hardware on a Stratix V 5SGXEA7K2F40C2N device. The operating frequency is 50 MHz and the average number of clocks for processing a set of four frames of QVGA images is 394k clocks. The proposed hardware achieved 520 fps, and can process multi-stream video from 64 cameras. The execution time on 3.4 GHz Core i7-3770 CPU was 8.4 fps. Therefore, the proposed hardware was about 62 times faster than that CPU.

This paper proposes an approach for the temporal segmentation of human body movements using IMU (Inertial Measurement Unit). The approach is based on online HMM-based segmentation of continuous time series data. In previous studies, the real-time segmentation of human body movement using joint angles acquired by optical motion capture has been realized, using stochastic motion modeling. The approach is now adapted for angular velocity data. The segmented motions are recognized via HMM models. The segmentation and recognition results of the proposed algorithm are demonstrated with experiments. Auto segmentation of each motion and recognition of motion patterns are verified using angular velocity data obtained by IMU sensors and the Wii remote. The success rate of auto segmentation using the data obtained by Wii remote was more than 80% on average.

Real-time object detection is important for surveillance applications. This paper describes a high-performance object detector using a commercially available FPGA. Major bottlenecks in the real AdaBoost classifier are resolved. A new FIR-filter-like hardware architecture takes advantage of an FPGA's hardware parallelism and block-RAM structure. The resulting design uses Xilinx Virtex 5 and achieves the real-time processing performance of 220 f/s at 201 MHz and adjustable recognition performance with a variable number of weak classifiers. This is the first demonstration of a histogram of oriented gradients and Real AdaBoost detector on an FPGA.

Network-on-chip (NoC) architectures are emerged as a promising solution to the lack of scalability in multiprocessor systems-on-chips (MPSoCs). In this paper, we propose a novel BMNoC configuration algorithm together with simulation results. Our BMNoC configuration algorithm analyses the data traffic of the target application and determines which core is the right one to put into the certain cluster with its communication volume and locality. Furthermore, the simulation results illustrate the better latency than earlier studies and feasibility of BMNoC.

We introduce a heuristic that generates as few a number of test vectors as possible with high branch coverage for the functional verification of digital design. The challenge is how to save time and effort for sufficient verification at system-level. We focus on generating test vectors from the circuit specification written in C. We reuse them to SystemC description by removing their redundancies while maintaining the branch coverage as is. Experimental results of our practical design show that over 90% on average of the redundant test vectors were reduced with 100% branch coverage maintained. The reused test vectors for SystemC Bus Cycle Accurate models scored 80% branch coverage on average. These results are significant for saving verification cost and beneficial for simplifying debugging works.

We present a novel approach for constructing a virtual 3D world from a sparse set of 2D photograph images. In our approach, we do not construct a large 3D world directly from the images, instead we construct several 3D models from the images and inter-connect them. Each 3D model is photograph-based 3D model that constitutes a few faces and a high-resolution photographic texture image. By arranging some photograph-based 3D models in a 3D scene, we construct a virtual 3D world. To inter-connect these 3D models and represent a unified 3D world, we render the 3D models by blending them together according to view-point position and rotation. Using our novel approach, it is possible to semi-automatically construct a virtual 3D world from fewer photograph images.

No abstract available.

To study autonomous control of a space robot, the authors developed their Advanced Space Robot Testbed with Redundant Arms (ASTRA). Berthing a satellite moving in space is a difficult and dangerous task when done manually. This paper describes a robot which berthed a mockup satellite moving and rotating in zero gravity. Features of the robot include autonomous approach to a moving target satellite using visual satellite motion estimation, using real-time visual tracking control to track satellite handles with two robot arms, and grasping the satellite handles with minimum mechanical shock using a flexible wrist mechanism and impedance control. The authors ran their satellite berthing experiment on the ASTRA to check its performance.>