Future embedded system products, e.g. smart hand-held mobile terminals, will accommodate a large number of applications that will partly run sequentially and independently, partly concurrently and interacting on massively parallel computing platforms. Already for systems of moderate complexity, the design space will be huge and its exploration requires that the system architect is able to quickly evaluate the performances of candidate architectures and application mappings. The mainstream evaluation technique today is the system-level performance simulation of the applications and platforms using abstracted workload and processing capacity models, respectively. These virtual system models allow fast simulation of large systems at an early phase of development with reasonable modeling effort and time. The accuracy of the performance results is dependent on how closely the models used reflect the actual system. This paper presents a compiler based technique for automatic generation of workload models for performance simulation, while exploiting an overall approach and platform performance capacity models developed previously. The resulting workload models are experimented using x264 video and JPEG encoding application examples.
Three-part white blood cell differentials which are key to routine blood workups are typically performed in centralized laboratories on conventional hematology analyzers operated by highly trained staff. With the trend of developing miniaturized blood analysis tool for point-of-need in order to accelerate turnaround times and move routine blood testing away from centralized facilities on the rise, our group has developed a highly miniaturized holographic imaging system for generating lens-free images of white blood cells in suspension. Analysis and classification of its output data, constitutes the final crucial step ensuring appropriate accuracy of the system. In this work, we implement reference holographic images of single white blood cells in suspension, in order to establish an accurate ground truth to increase classification accuracy. We also automate the entire workflow for analyzing the output and demonstrate clear improvement in the accuracy of the 3-part classification. High-dimensional optical and morphological features are extracted from reconstructed digital holograms of single cells using the ground-truth images and advanced machine learning algorithms are investigated and implemented to obtain 99% classification accuracy. Representative features of the three white blood cell subtypes are selected and give comparable results, with a focus on rapid cell recognition and decreased computational cost.
The presented platform-based object-oriented modeling concept for system design allowed us to create a networked hardware reconfigurable camera in a 25 man-month schedule with concurrent development of application and target FPGA platform. The developed TCP/IP layer achieves throughput of 2 Mb/s/MHz and the complete application logic consumes 700 mW at 20 MHz.
In this demo, we demonstrate a real-time viewpoint interpolation application on FPGA. Viewpoint interpolation is the process of synthesizing plausible in-between views - so-called virtual camera views - from a couple of surrounding fixed camera views. Stereo matching is used to extract depth information, by computing a disparity map from a pair of input images. With the depth information, virtual views at any points between the two cameras are computed through view interpolation. To make viewpoint interpolation possible for low/moderate-power consumer applications, a further quality/complexity tradeoff study is required to conciliate algorithmic quality to architectural performance. In essence, the inter-dependencies between the different algorithmic steps in the processing chain are thoroughly analyzed, aiming at an overall quality-performance model that pinpoints which algorithmic functionalities can be simplified with minor (preferably no) global input-output quality degradation, while maximally reducing their implementation complexity w.r.t. arithmetic and line buffer requirements. Compared to state-of-the-art CPU and GPU platforms running at several GHz clock speed, our low-power 65 MHz FPGA implementation achieves speedups with one order of magnitude over state-of-the-art, without impeding on the visual quality, reaching over 60 frames per second high-definition (1024×768) high-quality, 64-disparity search range stereo matching and enabling viewpoint interpolation in low-power, embedded applications.
Cardiotoxicity is the major cause of drug withdrawal from the market, despite rigorous toxicity testing during the drug development process. Existing safety screening techniques, some of which are based on simplified cellular assays, others on electrical (impedance) or optical (fluorescent microscopy) measurements, are either too limited in throughput or offer too poor predictability of toxicity to be applied on large numbers of compounds in the early stage of drug development. We present a compact optical system for direct (label-free) monitoring of fast cellular movements that enable low cost and high throughput drug screening. Our system is based on a high-speed lens-free in-line holographic microscope. When compared to a conventional microscope, the system can combine adequate imaging resolution (5.5 μm pixel pitch) with a large field-of-view (63.4 mm2) and high speed (170 fps) to capture physical cell motion in real-time. This combination enables registration of cardiac contractility parameters such as cell contraction frequency, total duration, and rate and duration of both contraction and relaxation. The system also quantifies conduction velocity, which is challenging in existing techniques. Additionally, to complement the imaging hardware we have developed image processing software that extracts all the contractility parameters directly from the raw interference images. The system was tested with varying concentration of the drug verapamil and at 100 nM, showed a decrease in: contraction frequency (-23.3% ± 13%), total duration (-21% ± 5%), contraction duration (-19% ± 6%) and relaxation duration (-21% ± 8%). Moreover, contraction displacement ceased at higher concentrations.
