Communications and High-Precision Positioning (CHP2): Hardware Architecture, Implementation, and Validation.

Spectral congestion and modern consumer applications motivate radio technologies that efficiently cooperate with nearby users and provide several services simultaneously. We designed and implemented a joint positioning-communications system that simultaneously enables network communications, timing synchronization, and localization to a variety of airborne and ground-based platforms. This Communications and High-Precision Positioning (CHP2) system simultaneously performs communications and precise ranging (<10 cm) with a narrow band waveform (10 MHz) at a carrier frequency of 915 MHz (US ISM) or 783 MHz (EU Licensed). The ranging capability may be extended to estimate the relative position and orientation by leveraging the spatial diversity of the multiple-input, multiple-output (MIMO) platforms. CHP2 also digitally synchronizes distributed platforms with sub-nanosecond precision without support from external systems (GNSS, GPS, etc.). This performance is enabled by leveraging precise time-of-arrival (ToA) estimation techniques, a network synchronization algorithm, and the intrinsic cooperation in the joint processing chain that executes these tasks simultaneously. In this manuscript, we describe the CHP2 system architecture, hardware implementation, and in-lab and over-the-air experimental validation.

T-BFA: Targeted Bit-Flip Adversarial Weight Attack.

Traditional Deep Neural Network (DNN) security is mostly related to the well-known adversarial input example attack.Recently, another dimension of adversarial attack, namely, attack on DNN weight parameters, has been shown to be very powerful. Asa representative one, the Bit-Flip based adversarial weight Attack (BFA) injects an extremely small amount of faults into weight parameters to hijack the executing DNN function. Prior works of BFA focus on un-targeted attacks that can hack all inputs into a random output class by flipping a very small number of weight bits stored in computer memory. This paper proposes the first work oftargetedBFA based (T-BFA) adversarial weight attack on DNNs, which can intentionally mislead selected inputs to a target output class. The objective is achieved by identifying the weight bits that are highly associated with classification of a targeted output through a class-dependent weight bit searching algorithm. Our proposed T-BFA performance is successfully demonstrated on multiple DNN architectures for image classification tasks. For example, by merely flipping 27 out of 88 million weight bits of ResNet-18, our T-BFA can misclassify all the images from Hen class into Goose class (i.e., 100% attack success rate) in ImageNet dataset, while maintaining 59.35% validation accuracy.

Articulation constrained learning with application to speech emotion recognition.

Speech emotion recognition methods combining articulatory information with acoustic features have been previously shown to improve recognition performance. Collection of articulatory data on a large scale may not be feasible in many scenarios, thus restricting the scope and applicability of such methods. In this paper, a discriminative learning method for emotion recognition using both articulatory and acoustic information is proposed. A traditional ℓ 1-regularized logistic regression cost function is extended to include additional constraints that enforce the model to reconstruct articulatory data. This leads to sparse and interpretable representations jointly optimized for both tasks simultaneously. Furthermore, the model only requires articulatory features during training; only speech features are required for inference on out-of-sample data. Experiments are conducted to evaluate emotion recognition performance over vowels /AA/,/AE/,/IY/,/UW/ and complete utterances. Incorporating articulatory information is shown to significantly improve the performance for valence-based classification. Results obtained for within-corpus and cross-corpus categorical emotion recognition indicate that the proposed method is more effective at distinguishing happiness from other emotions.

High-Volume-Rate 3-D Ultrasound Imaging Based on Synthetic Aperture Sequential Beamforming With Chirp-Coded Excitation.

Three-dimensional (3-D) ultrasound imaging is a promising modality for many medical applications. Unfortunately, it generates voluminous data in the front end, making it unattractive for high-volume-rate portable medical applications. We apply synthetic aperture sequential beamforming (SASB) to greatly compress the front-end receive data. Baseline 3-D SASB has a low volume rate, because subapertures fire one by one. In this paper, we propose to increase the volume rate of 3-D SASB without degrading imaging quality through: 1) transmitting and receiving simultaneously with four subapertures and 2) using linear chirps as the excitation waveform to reduce interference. We design four linear chirps that operate on two overlapped frequency bands with chirp pairs in each band having opposite chirp rates. Direct implementation of this firing scheme results in grating lobes. Therefore, we design a sparse array that mitigates the grating lobe levels through optimizing the locations of transducer elements in the bin-based random array. Compared with the baseline 3-D SASB, the proposed method increases the volume rate from 8.56 to 34.2 volumes/s without increasing the front-end computation requirement. Field-II-based cyst simulations show that the proposed method achieves imaging quality comparable with baseline 3-D SASB in both shallow and deep regions.

Error analysis of speed of sound reconstruction in ultrasound limited angle transmission tomography.

We have investigated limited angle transmission tomography to estimate speed of sound (SOS) distributions for breast cancer detection. That requires both accurate delineations of major tissues, in this case by segmentation of prior B-mode images, and calibration of the relative positions of the opposed transducers. Experimental sensitivity evaluation of the reconstructions with respect to segmentation and calibration errors is difficult with our current system. Therefore, parametric studies of SOS errors in our bent-ray reconstructions were simulated. They included mis-segmentation of an object of interest or a nearby object, and miscalibration of relative transducer positions in 3D. Close correspondence of reconstruction accuracy was verified in the simplest case, a cylindrical object in homogeneous background with induced segmentation and calibration inaccuracies. Simulated mis-segmentation in object size and lateral location produced maximum SOS errors of 6.3% within 10 mm diameter change and 9.1% within 5 mm shift, respectively. Modest errors in assumed transducer separation produced the maximum SOS error from miscalibrations (57.3% within 5 mm shift), still, correction of this type of error can easily be achieved in the clinic. This study should aid in designing adequate transducer mounts and calibration procedures, and in specification of B-mode image quality and segmentation algorithms for limited angle transmission tomography relying on ray tracing algorithms.

Low-Cost 3-D Flow Estimation of Blood With Clutter.

Volumetric flow rate estimation is an important ultrasound medical imaging modality that is used for diagnosing cardiovascular diseases. Flow rates are obtained by integrating velocity estimates over a cross-sectional plane. Speckle tracking is a promising approach that overcomes the angle dependency of traditional Doppler methods, but suffers from poor lateral resolution. Recent work improves lateral velocity estimation accuracy by reconstructing a synthetic lateral phase (SLP) signal. However, the estimation accuracy of such approaches is compromised by the presence of clutter. Eigen-based clutter filtering has been shown to be effective in removing the clutter signal; but it is computationally expensive, precluding its use at high volume rates. In this paper, we propose low-complexity schemes for both velocity estimation and clutter filtering. We use a two-tiered motion estimation scheme to combine the low complexity sum-of-absolute-difference and SLP methods to achieve subpixel lateral accuracy. We reduce the complexity of eigen-based clutter filtering by processing in subgroups and replacing singular value decomposition with less compute-intensive power iteration and subspace iteration methods. Finally, to improve flow rate estimation accuracy, we use kernel power weighting when integrating the velocity estimates. We evaluate our method for fast- and slow-moving clutter for beam-to-flow angles of 90° and 60° using Field II simulations, demonstrating high estimation accuracy across scenarios. For instance, for a beam-to-flow angle of 90° and fast-moving clutter, our estimation method provides a bias of -8.8% and standard deviation of 3.1% relative to the actual flow rate.

A distributed Canny edge detector: algorithm and FPGA implementation.

The Canny edge detector is one of the most widely used edge detection algorithms due to its superior performance. Unfortunately, not only is it computationally more intensive as compared with other edge detection algorithms, but it also has a higher latency because it is based on frame-level statistics. In this paper, we propose a mechanism to implement the Canny algorithm at the block level without any loss in edge detection performance compared with the original frame-level Canny algorithm. Directly applying the original Canny algorithm at the block-level leads to excessive edges in smooth regions and to loss of significant edges in high-detailed regions since the original Canny computes the high and low thresholds based on the frame-level statistics. To solve this problem, we present a distributed Canny edge detection algorithm that adaptively computes the edge detection thresholds based on the block type and the local distribution of the gradients in the image block. In addition, the new algorithm uses a nonuniform gradient magnitude histogram to compute block-based hysteresis thresholds. The resulting block-based algorithm has a significantly reduced latency and can be easily integrated with other block-based image codecs. It is capable of supporting fast edge detection of images and videos with high resolutions, including full-HD since the latency is now a function of the block size instead of the frame size. In addition, quantitative conformance evaluations and subjective tests show that the edge detection performance of the proposed algorithm is better than the original frame-based algorithm, especially when noise is present in the images. Finally, this algorithm is implemented using a 32 computing engine architecture and is synthesized on the Xilinx Virtex-5 FPGA. The synthesized architecture takes only 0.721 ms (including the SRAM READ/WRITE time and the computation time) to detect edges of 512 × 512 images in the USC SIPI database when clocked at 100 MHz and is faster than existing FPGA and GPU implementations.