Runtime Performance Research Articles

Vehicle detection using improved region convolution neural network for accident prevention in smart roads

• We develop a novel cleaning algorithm to remove the unnecessary features based on SIFT extractor. • An accurate object detection model is proposed by adopting the G-RCNN to process vehicle image data. • We develop an evolutionary algorithm to intelligently explore the hyper-parameters space of the IRCNN-VD. This paper explores the vehicle detection problem and introduces an improved regional convolution neural network. The vehicle data (set of images) is first collected, from which the noise (set of outlier images) is removed using the SIFT extractor. The region convolution neural network is then used to detect the vehicles. We propose a new hyper-parameters optimization model based on evolutionary computation that can be used to tune parameters of the deep learning framework. The proposed solution was tested using the well-known boxy vehicle detection data , which contains more than 200,000 vehicle images and 1,990,000 annotated vehicles. The results are very promising and show superiority over many current state-of-the-art solutions in terms of runtime and accuracy performances.

Coarse Grained FPGA Overlay for Rapid Just-In-Time Accelerator Compilation

Coarse-grained FPGA overlays built around the runtime programmable DSP blocks in modern FPGAs can achieve high throughput and improved scalability compared to traditional overlays built without detailed consideration of FPGA architecture. These overlays can be mapped to using higher level compilers, achieving fast compilation, software-like programmability and run-time management, and high-level design abstraction. OpenCL allows programs running on a host computer to launch accelerator kernels which can be compiled at run-time for a specific architecture, thus enabling portability. However, prohibitive hardware compilation times in traditional design flows mean that the tools cannot effectively use just-in-time (JIT) compilation or runtime performance scaling on FPGAs. We present a methodology for runtime compilation of dataflow graphs expressed as OpenCL kernels onto coarse-grained overlays. The methodology benefits from the high level of abstraction afforded by using the OpenCL programming model, while the mapping to the overlay significantly reduces compilation and load times. Key characteristics of this work include highly performant DSP-optimized functional units that scale to large overlays on modern devices and the ability to perform automatic resource-aware kernel replication up to the size of the overlay. We demonstrate place and route times orders of magnitude better than traditional HLS flows, even when running on an embedded processor in the Xilinx Zynq.

Collaborative Cluster Configuration for Distributed Data-Parallel Processing: A Research Overview

Many organizations routinely analyze large datasets using systems for distributed data-parallel processing and clusters of commodity resources. Yet, users need to configure adequate resources for their data processing jobs. This requires significant insights into expected job runtimes and scaling behavior, resource characteristics, input data distributions, and other factors. Unable to estimate performance accurately, users frequently overprovision resources for their jobs, leading to low resource utilization and high costs.In this paper, we present major building blocks towards a collaborative approach for optimization of data processing cluster configurations based on runtime data and performance models. We believe that runtime data can be shared and used for performance models across different execution contexts, significantly reducing the reliance on the recurrence of individual processing jobs or, else, dedicated job profiling. For this, we describe how the similarity of processing jobs and cluster infrastructures can be employed to combine suitable data points from local and global job executions into accurate performance models. Furthermore, we outline approaches to performance prediction via more context-aware and reusable models. Finally, we lay out how metrics from previous executions can be combined with runtime monitoring to effectively re-configure models and clusters dynamically.

Mapping and Validating a Point Neuron Model on Intel's Neuromorphic Hardware Loihi.

Neuromorphic hardware is based on emulating the natural biological structure of the brain. Since its computational model is similar to standard neural models, it could serve as a computational accelerator for research projects in the field of neuroscience and artificial intelligence, including biomedical applications. However, in order to exploit this new generation of computer chips, we ought to perform rigorous simulation and consequent validation of neuromorphic models against their conventional implementations. In this work, we lay out the numeric groundwork to enable a comparison between neuromorphic and conventional platforms. “Loihi”—Intel's fifth generation neuromorphic chip, which is based on the idea of Spiking Neural Networks (SNNs) emulating the activity of neurons in the brain, serves as our neuromorphic platform. The work here focuses on Leaky Integrate and Fire (LIF) models based on neurons in the mouse primary visual cortex and matched to a rich data set of anatomical, physiological and behavioral constraints. Simulations on classical hardware serve as the validation platform for the neuromorphic implementation. We find that Loihi replicates classical simulations very efficiently with high precision. As a by-product, we also investigate Loihi's potential in terms of scalability and performance and find that it scales notably well in terms of run-time performance as the simulated networks become larger.

Mapping and Validating a Point Neuron Model on Intel's Neuromorphic Hardware Loihi.

Neuromorphic hardware is based on emulating the natural biological structure of the brain. Since its computational model is similar to standard neural models, it could serve as a computational accelerator for research projects in the field of neuroscience and artificial intelligence, including biomedical applications. However, in order to exploit this new generation of computer chips, we ought to perform rigorous simulation and consequent validation of neuromorphic models against their conventional implementations. In this work, we lay out the numeric groundwork to enable a comparison between neuromorphic and conventional platforms. "Loihi"-Intel's fifth generation neuromorphic chip, which is based on the idea of Spiking Neural Networks (SNNs) emulating the activity of neurons in the brain, serves as our neuromorphic platform. The work here focuses on Leaky Integrate and Fire (LIF) models based on neurons in the mouse primary visual cortex and matched to a rich data set of anatomical, physiological and behavioral constraints. Simulations on classical hardware serve as the validation platform for the neuromorphic implementation. We find that Loihi replicates classical simulations very efficiently with high precision. As a by-product, we also investigate Loihi's potential in terms of scalability and performance and find that it scales notably well in terms of run-time performance as the simulated networks become larger.

A trust-region augmented Hessian implementation for state-specific and state-averaged CASSCF wave functions.

In this work, we present a one-step second-order converger for state-specific (SS) and state-averaged (SA) complete active space self-consistent field (CASSCF) wave functions. Robust convergence is achieved through step restrictions using a trust-region augmented Hessian (TRAH) algorithm. To avoid numerical instabilities, an exponential parameterization of variational configuration parameters is employed, which works with a nonredundant orthogonal complement basis. This is a common approach for SS-CASSCF and is extended to SA-CASSCF wave functions in this work. Our implementation is integral direct and based on intermediates that are formulated in either the sparse atomic-orbital or small active molecular-orbital basis. Thus, it benefits from a combination with efficient integral decomposition techniques, such as the resolution-of-the-identity or the chain-of-spheres for exchange approximations. This facilitates calculations on large molecules, such as a Ni(II) complex with 231 atoms and 5154 basis functions. The runtime performance of TRAH-CASSCF is competitive with the other state-of-the-art implementations of approximate and full second-order algorithms. In comparison with a sophisticated first-order converger, TRAH-CASSCF calculations usually take more iterations to reach convergence and, thus, have longer runtimes. However, TRAH-CASSCF calculations still converge reliably to a true minimum even if the first-order algorithm fails.

PyCRTM: A python interface for the community radiative transfer model

The Community Radiative Transfer Model (CRTM) is a powerful and versatile scalar radiative transfer model for satellite data assimilation and remote sensing applications. It is implemented as an object-oriented Fortran library, enabling flexible code development and optimal runtime performance on clusters. The downsides of the Fortran interface are a steep learning curve for students and the reduced productivity of users that is typical for static compiled languages, in contrast to dynamic interpreted languages like Python. pyCRTM is a new software framework that directly interfaces the CRTM Fortran data structures and procedures in Python, leveraging both the simplicity and ease of use of Python syntax as well as the flexibility arising from the vast contemporary Python ecosystem. The goal of pyCRTM is to lower the barrier of entry for university students to learn and use the CRTM and to boost the productivity of researchers seeking to create new methods in radiative transfer and data assimilation, or seeking to apply the CRTM to study atmospheric phenomena without having to go through the pre-existing complexity of the CRTM Fortran interface.

SVBRDF Recovery from a Single Image with Highlights Using a Pre‐trained Generative Adversarial Network

AbstractSpatially varying bi‐directional reflectance distribution functions (SVBRDFs) are crucial for designers to incorporate new materials in virtual scenes, making them look more realistic. Reconstruction of SVBRDFs is a long‐standing problem. Existing methods either rely on an extensive acquisition system or require huge datasets, which are non‐trivial to acquire. We aim to recover SVBRDFs from a single image, without any datasets. A single image contains incomplete information about the SVBRDF, making the reconstruction task highly ill‐posed. It is also difficult to separate between the changes in colour that are caused by the material and those caused by the illumination, without the prior knowledge learned from the dataset. In this paper, we use an unsupervised generative adversarial neural network (GAN) to recover SVBRDFs maps with a single image as input. To better separate the effects due to illumination from the effects due to the material, we add the hypothesis that the material is stationary and introduce a new loss function based on Fourier coefficients to enforce this stationarity. For efficiency, we train the network in two stages: reusing a trained model to initialize the SVBRDFs and fine‐tune it based on the input image. Our method generates high‐quality SVBRDFs maps from a single input photograph, and provides more vivid rendering results compared to the previous work. The two‐stage training boosts runtime performance, making it eight times faster than the previous work.

Inference-aware convolutional neural network pruning

Deep neural networks (DNNs) have become an important tool in solving various problems in numerous disciplines. However, DNNs are also known for their high resource requirement, weight redundancy, and large-scale parameters. As a result, the use of DNNs is restricted for devices that lack the necessary resources required to execute, especially resource-constrained devices such as mobile phones, wearable devices, and other edge devices. In recent years, pruning has emerged as an essential technique to reduce insignificant parameters and accelerate the model performance. However, finding the optimal number of parameters that can be pruned without significantly affecting the model performance is a time-consuming, tedious task and require a lot of manual tuning. This paper represent pruning as an optimization problem with the goal of improving DNN run-time inference performance by pruning low impacting parameters (filters) and their corresponding feature maps. To do this, we present a Bayesian optimization-based method for automatically determining the appropriate number of filters for each convolutional layer. Also, we proposed an objective function incorporating distinct model performance and resource-specific constraints. The proposed method is applied to two different kinds of convolutional network architectures (i.e., VGG16 and deeper network ResNet34) on CIFAR10, CIFAR100, and ImageNet datasets. The large-scale ImageNet experimental findings showed that the floating-point operations of the ResNet34 and VGG16 could be reduced by 35.46 percent and 84.97 percent, respectively, with negligible loss of accuracy.

Fast and Efficient Data Masking Method for Securing Image Over Cloud Computing Environment

Objective: Preserving the confidentiality of sensitive information is becoming more and more difficult and challenging, considering current scenarios, as a huge amount of multimedia data are stored and communicated over the internet among users and cloud computing environment. The existing cryptography security model for storing images on a cloud platform cannot resist various kinds of modern attacks, such as statistical, differential, brute force, cropping attack, etc., therefore, an improved bit scrambling technique using chaotic maps that can resist various kinds of security attacks is needed. The FEDM cipher image provides less correlation among neighboring pixels and images can be decrypted even under the presence of noise. This study proposed a FEDM model to achieve better UACI, NPCR, histogram, runtime, and processing time performance than the existing image security methods. Methods: Preserving the confidentiality of sensitive information is becoming more and more difficult and challenging considering current scenarios as a huge amount of multimedia data are stored and communicated over the internet among users and cloud computing environment. The existing cryptography security model for storing images on a cloud platform cannot resist various kinds of modern attacks such as statistical, differential, brute force, cropping attack, etc. Results: The overall results show that the proposed FEDM model attains much superior performance considering histogram, UACI, NPCR, and runtime. The FEDM model can resist against SA. The FEDM model attains better performance because IBS is used in each step of CS. Thus, a correlation between adjacent pixels is less and aids superior security performance. Further, the FEDM model attains better UACI and NPCR performance when compared with the exiting image encryption model. Conclusion: The FEDM security method can resist DA, noise, cropping attack, and linear attacks more efficiently due to a larger keyspace. Further, the FEDM takes less time for provisioning security. Along with this, FEDM works smoothly under a cloud computing environment. No prior work has considered runtime performance evaluation under the cloud computing environment. FEDM model will significantly aid in reducing the overall operational cost of a cloud computing environment with a reduction in processing time as cloud charge is based on hours of usage.

Complexity-guided container replacement synthesis

Containers, such as lists and maps, are fundamental data structures in modern programming languages. However, improper choice of container types may lead to significant performance issues. This paper presents Cres, an approach that automatically synthesizes container replacements to improve runtime performance. The synthesis algorithm works with static analysis techniques to identify how containers are utilized in the program, and attempts to select a method with lower time complexity for each container method call. Our approach can preserve program behavior and seize the opportunity of reducing execution time effectively for general inputs. We implement Cres and evaluate it on 12 real-world Java projects. It is shown that Cres synthesizes container replacements for the projects with 384.2 KLoC in 14 minutes and discovers six categories of container replacements, which can achieve an average performance improvement of 8.1%.

Burapha-TH: A Multi-Purpose Character, Digit, and Syllable Handwriting Dataset

In handwriting recognition research, a public image dataset is necessary to evaluate algorithm correctness and runtime performance. Unfortunately, in existing Thai language script image datasets, there is a lack of variety of standard handwriting types. This paper focuses on a new offline Thai handwriting image dataset named Burapha-TH. The dataset has 68 character classes, 10 digit classes, and 320 syllable classes. For constructing the dataset, 1072 Thai native speakers wrote on collection datasheets that were then digitized using a 300 dpi scanner. De-skewing, detection box and segmentation algorithms were applied to the raw scans for image extraction. The experiment used different deep convolutional models with the proposed dataset. The result shows that the VGG-13 model (with batch normalization) achieved accuracy rates of 95.00%, 98.29%, and 96.16% on character, digit, and syllable classes, respectively. The Burapha-TH dataset, unlike all other known Thai handwriting datasets, retains existing noise, the white background, and all artifacts generated by scanning. This comprehensive, raw, and more realistic dataset will be helpful for a variety of research purposes in the future.

Philympics 2021: Prophage Predictions Perplex Programs

Background Most bacterial genomes contain integrated bacteriophages—prophages—in various states of decay. Many are active and able to excise from the genome and replicate, while others are cryptic prophages, remnants of their former selves. Over the last two decades, many computational tools have been developed to identify the prophage components of bacterial genomes, and it is a particularly active area for the application of machine learning approaches. However, progress is hindered and comparisons thwarted because there are no manually curated bacterial genomes that can be used to test new prophage prediction algorithms. Methods We present a library of gold-standard bacterial genomes with manually curated prophage annotations, and a computational framework to compare the predictions from different algorithms. We use this suite to compare all extant stand-alone prophage prediction algorithms and identify their strengths and weaknesses. We provide a FAIR dataset for prophage identification, and demonstrate the accuracy, precision, recall, and f 1 score from the analysis of ten different algorithms for the prediction of prophages. Results We identified strengths and weaknesses between the prophage prediction tools. Several tools exhibit exceptional f 1 scores, while others have better recall at the expense of more false positives. The tools vary greatly in runtime performance with few exhibiting all desirable qualities for large-scale analyses. Conclusions Our library of gold-standard prophage annotations and benchmarking framework provide a valuable resource for exploring strengths and weaknesses of current and future prophage annotation tools. We discuss caveats and concerns in this analysis, how those concerns may be mitigated, and avenues for future improvements. This framework will help developers identify opportunities for improvement and test updates. It will also help users in determining the tools that are best suited for their analysis.

Tofu: a fast, versatile and user-friendly image processing toolkit for computed tomography.

Tofu is a toolkit for processing large amounts of images and for tomographic reconstruction. Complex image processing tasks are organized as workflows of individual processing steps. The toolkit is able to reconstruct parallel and cone beam as well as tomographic and laminographic geometries. Many pre- and post-processing algorithms needed for high-quality 3D reconstruction are available, e.g. phase retrieval, ring removal and de-noising. Tofu is optimized for stand-alone GPU workstations on which it achieves reconstruction speed comparable with costly CPU clusters. It automatically utilizes all GPUs in the system and generates 3D reconstruction code with minimal number of instructions given the input geometry (parallel/cone beam, tomography/laminography), hence yielding optimal run-time performance. In order to improve accessibility for researchers with no previous knowledge of programming, tofu contains graphical user interfaces for both optimization of 3D reconstruction parameters and batch processing of data with pre-configured workflows for typical computed tomography reconstruction. The toolkit is open source and extensive documentation is available for both end-users and developers. Thanks to the mentioned features, tofu is suitable for both expert users with specialized image processing needs (e.g. when dealing with data from custom-built computed tomography scanners) and for application-specific end-users who just need to reconstruct their data on off-the-shelf hardware.

Resilient Back Propagation Neural Network Security Model For Containerized Cloud Computing

Cloud-native computing is getting more and more popular in recent years where containerized microservices architectural designs play a central role in building a distributed systems and services. On one hand, they bring convenience and simplicity to build massively scalable distributed cloud-native applications and enable continuous development and delivery for their services. On the other hand, they widen the surface of malicious intrusions, which, in turn, without proper defense mechanisms, lessens their benefits to a certain degree. Among the biggest threats of malicious intrusions are those that belong to the Distributed Denial of Service (DDoS) family. Such type of attacks are challenging because DDoS attacks are elevated hard-to-absorbed threats and have a high degree of variability in types, design, and complexity. In this work, resilient backpropagation neural network was used to build an intelligent network intrusion detection model against the most modernistic DDoS attacks in the cloud-native computing environment. We evaluated our proposed model using the benchmarking Canadian Institute for Cybersecurity evaluation CICDDoS 2019 dataset. Our proposed detection model has achieved high reflective DDoS attack detection. Therefore, it is appropriate to defend against reflective DDoS attacks in containerized cloud-native platforms. Experimental results indicate that the DDoS attack detection accuracy of the proposed resilient neural network model is as high as 97.07% which outperforms most of the well-known learning models mentioned in the most related work. Moreover, the proposed model has achieved a competitive run time performance that highly meets the delay requirements of containerized cloud computing.

Anytime 3D Object Reconstruction Using Multi-Modal Variational Autoencoder

For effective human-robot teaming, it is important for the robots to be able to share their visual perception with the human operators. In a harsh remote collaboration setting, data compression techniques such as autoencoder can be utilized to obtain and transmit the data in terms of latent variables in a compact form. In addition, to ensure real-time runtime performance even under unstable environments, an anytime estimation approach is desired that can reconstruct the full contents from incomplete information. In this context, we propose a method for imputation of latent variables whose elements are partially lost. To achieve the anytime property with only a few dimensions of variables, exploiting prior information of the category-level is essential. A prior distribution used in variational autoencoders is simply assumed to be isotropic Gaussian regardless of the labels of each training datapoint. This type of flattened prior makes it difficult to perform imputation from the category-level distributions. We overcome this limitation by exploiting a category-specific multi-modal prior distribution in the latent space. The missing elements of the partially transferred data can be sampled, by finding a specific modal according to the remaining elements. Since the method is designed to use partial elements for anytime estimation, it can also be applied for data over-compression. Based on the experiments on the ModelNet and Pascal3D datasets, the proposed approach shows consistently superior performance over autoencoder and variational autoencoder up to 70% data loss. The software is open source and is available from our repository <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> .

Time-Optimized Online Planning For Parallel Parking With Nonlinear Optimization and Improved Monte Carlo Tree Search

Automatic parallel parking is critical to increase safety in urban narrow parking spots, maximize the traffic efficiency, and provide human drivers with mobility and convenience. Recent research integrates Monte Carlo tree search (MCTS) and artificial neural networks (ANNs) to calculate optimal lateral motions without considering the longitudinal aspect and narrow spots; advances in nonlinear programming-based (NPB) parking methods consider time-optimal parking motion in narrow spots using the time-consuming optimization calculation. To address the computational efficiency of the planning of time-optimal parking maneuvers, a complete framework relying on the two compositions was introduced. First, nonlinear optimization that formulates the minimum motion time and vehicle constraint was used to generate the data of parking motions offline. These motions were subsequently learned by ANNs. Second, the ANNs trained on the offline data were employed by an improved MCTS to generate approximate time-optimal parking motions online. The time-optimized performance and run-time performance of the proposed method were confirmed by comparing with that of NPB and other mainstream methods. A success rate of 100% for parking slots with merely 10% larger length than the vehicle was realized in simulations. Experiments were conducted on a full-sized pure electric vehicle to further confirm the effectiveness of the proposed method.

Experimental Comparisons of Clustering Approaches for Data Representation

Clustering approaches are extensively used by many areas such as IR, Data Integration, Document Classification, Web Mining, Query Processing, and many other domains and disciplines. Nowadays, much literature describes clustering algorithms on multivariate data sets. However, there is limited literature that presented them with exhaustive and extensive theoretical analysis as well as experimental comparisons. This experimental survey paper deals with the basic principle, and techniques used, including important characteristics, application areas, run-time performance, internal, external, and stability validity of cluster quality, etc., on five different data sets of eleven clustering algorithms. This paper analyses how these algorithms behave with five different multivariate data sets in data representation. To answer this question, we compared the efficiency of eleven clustering approaches on five different data sets using three validity metrics-internal, external, and stability and found the optimal score to know the feasible solution of each algorithm. In addition, we have also included four popular and modern clustering algorithms with only their theoretical discussion. Our experimental results for only traditional clustering algorithms showed that different algorithms performed different behavior on different data sets in terms of running time (speed), accuracy and, the size of data set. This study emphasized the need for more adaptive algorithms and a deliberate balance between the running time and accuracy with their theoretical as well as implementation aspects.

APE: Metrics for understanding application performance efficiency under power caps

As supercomputers continue to grow in size and power consumption, the ability to understand application run time performance and energy/power usage is of growing importance. In the future it may be necessary for systems to operate under power caps imposed by facilities due to external influences such as renewable power generation (e.g. solar) impacting energy availability at an affordable cost during certain times of day and more frequent natural disasters due to climate change causing power transmission disruptions. However, current energy consumption and power characterization metrics like energy-delay products do not express all of the characteristics of an application that are relevant to understanding the impact of power capping on performance.In this study, we (1) characterize the useful features of a metric that effectively captures time-to-solution and power performance dynamics; and (2) design and evaluate a novel set of application power efficiency (APE) metrics that accounts for time-to-completion, power utilization and power variability. Power utilization quantifies a system’s trapped capacity (unused power budget) and power variability quantifies an application’s variation of power consumption over time. We demonstrate how APE can be used to quantify application characteristics which, in turn, enables reasoning about the impacts of power capping on an application.

Machine learning-based early detection of IoT botnets using network-edge traffic

In this work, we present an IoT botnet detection solution, EDIMA, consisting of a set of lightweight modules designed to be deployed at the edge gateway installed in home networks with the remaining modules expected to be implemented on cloud servers. EDIMA targets early detection of IoT botnets prior to the launch of an attack and includes a novel two-stage Machine Learning (ML)-based detector developed specifically for IoT bot detection at the edge gateway. The ML-based bot detector first employs supervised ML algorithms for aggregate traffic classification and subsequently Autocorrelation Function (ACF)-based tests to detect individual bots. The EDIMA architecture also comprises a malware traffic database, a policy engine, a feature extractor and a traffic parser. Performance evaluation results using our testbed setup with real-world IoT malware traffic as well as other public IoT datasets show that EDIMA achieves high bot scanning and bot-CnC traffic detection accuracies with very low false positive rates. The detection performance is also shown to be robust to an increase in the number of IoT devices connected to the edge gateway where EDIMA is deployed. Further, the runtime performance analysis of a Python implementation of EDIMA deployed on a Raspberry Pi reveals low bot detection delays and low RAM consumption. EDIMA is also shown to outperform existing detection techniques for bot scanning traffic and bot-CnC server communication.