Intermediate Representation Research Articles

Roadmap to Reasoning in Microservice Systems: A Rapid Review

Understanding software systems written by others is often challenging. When we want to assess systems to reason about them, i.e., to understand dependencies, analyze evolution trade-offs, or to verify conformance to the original blueprint, we must invest broad efforts. This becomes difficult when considering decentralized systems. Microservice-based systems are mainstream these days; however, to observe, understand, and manage these systems and their properties, we are missing fundamental tools that would derive various simplified system abstract perspectives. Microservices architecture characteristics yield many advantages to system operation; however, they bring challenges to their development and deployment lifecycles. Microservices urge a system-centric perspective to better reason about the system evolution and its quality attributes. This process review paper considers the current system analysis approaches and their possible alignment with automated system assessment or with human-centered approaches. We outline the necessary steps to accomplish holistic reasoning in decentralized microservice systems. As a contribution, we provide a roadmap for analysis and reasoning in microservice-based systems and suggest that various process phases can be decoupled through the introduction of system intermediate representation as the trajectory to provide various system-centered perspectives to analyze various system aspects. Furthermore, we cover different technical-based reasoning strategies and metrics in addition to the human-centered reasoning addressed through alternative visualization approaches. Finally, a system evolution is discussed from the perspective of such a reasoning process to illustrate the impact analysis evaluation over system changes.

MSmart: Smart Contract Vulnerability Analysis and Improved Strategies Based on Smartcheck

As is well known, smart contracts on the blockchain store plenty of digital assets, and these contracts deployed on the blockchain are difficult to be modified. For this reason, the analysis and detection of smart contract vulnerabilities have received extensive attention. Smartcheck, a typical Java-implemented static analysis tool of smart contracts, is capable of converting Solidity source code into path diagrams based on the lexical and syntactic analysis, and finds smart contract vulnerabilities by path matching. Although Smartcheck can analyze most of the real-world vulnerabilities, some imperceptible vulnerabilities may be ignored, causing huge economic losses. In order to address these issues, we develop a new tool named MSmart to analyze the vulnerabilities of high risk such as timestamp dependence vulnerabilities, integer overflow vulnerabilities, self-destruct vulnerabilities, etc. MSmart converts the smart contract source code into an intermediate representation, and looks for smart contract vulnerabilities based on intermediate representation and XPath rules. We add new intermediate representation rules of Smartcheck to detect more kinds of vulnerabilities and optimize existing rules to suit the complexity of smart contract. We also implemented smart contract batch detection to shorten the time it takes to find vulnerabilities. To analysis the performance of MSmart, we collect 6000 real-world contracts from Etherscan and design some comparative experiments with other tools. The results of experiment show that MSmart is able to analyze related vulnerabilities better, and false positives and false negatives have been reduced due to our improvements.

ViT based Covid-19 detection and classification from CXR images

The COVID-19 virus induces infection in both the upper respiratory tract and the lungs. Chest X-ray are widely used to diagnose various lung diseases. Considering chest X-ray and CT images, we explore deep-learning-based models namely: AlexNet, VGG16, VGG19, Resnet50, and Resnet101v2 to classify images representing COVID-19 infection and normal health situation. We analyze and present the impact of transfer learning, normalization, resizing, augmentation, and shuffling on the performance of these models. We explored the vision transformer (ViT) model to classify the CXR images. The ViT model incorporates multi-headed attention to disclose more global information in constrast to CNN models at lower layers. This mechanism leads to quantitatively diverse features. The ViT model renders consolidated intermediate representations considering the training data. For experimental analysis, we use two standard datasets and exploit performance metrics: accuracy, precision, recall, and F1-score. The ViT model, driven by self-attention mechanism and longrange context learning, outperforms other models.

Taking Back Control in an Intermediate Representation for GPU Computing

We describe our experiences successfully applying lightweight formal methods to substantially improve and reformulate an important part of Standard Portable Intermediate Representation SPIRV, an industry-standard language for GPU computing. The formal model that we present has allowed us to (1) identify several ambiguities and needless complexities in the way that structured control flow was defined in the SPIRV specification; (2) interact with the authors of the SPIRV specification to rectify these problems; (3) validate the developer tools and conformance test suites that support the SPIRV language by cross-checking them against our formal model, improving the tools, test suites, and our models in the process; and (4) develop a novel method for fuzzing SPIRV compilers to detect miscompilation bugs that leverages our formal model. The latest release of the SPIRV specification incorporates the revised set of control-flow definitions that have arisen from our work. Furthermore, our novel compiler-fuzzing technique has led to the discovery of twenty distinct, previously unknown bugs in SPIRV compilers from Google, the Khronos Group, Intel, and Mozilla. Our work showcases the practical impact that formal modelling and analysis techniques can have on the design and implementation of industry-standard programming languages.

SSA Translation Is an Abstract Interpretation

Static single assignment (SSA) form is a popular intermediate representation that helps implement useful static analyses, including global value numbering (GVN), sparse dataflow analyses, or SMT-based abstract interpretation or model checking. However, the precision of the SSA translation itself depends on static analyses, and a priori static analysis is even indispensable in the case of low-level input languages like machine code.To solve this chicken-and-egg problem, we propose to turn the SSA translation into a standard static analysis based on abstract interpretation. This allows the SSA translation to be combined with other static analyses in a single pass, taking advantage of the fact that it is more precise to combine analyses than applying passes in sequence.We illustrate the practicality of these results by writing a simple dataflow analysis that performs SSA translation, optimistic global value numbering, sparse conditional constant propagation, and loop-invariant code motion in a single small pass; and by presenting a multi-language static analyzer for both C and machine code that uses the SSA abstract domain as its main intermediate representation.

Multi-view Shape Generation for a 3D Human-like Body

Three-dimensional (3D) human-like body reconstruction via a single RGB image has attracted significant research attention recently. Most of the existing methods rely on the Skinned Multi-Person Linear model and thus can only predict unified human bodies. Moreover, meshes reconstructed by current methods sometimes perform well from a canonical view but not from other views, as the reconstruction process is commonly supervised by only a single view. To address these limitations, this article proposes a multi-view shape generation network for a 3D human-like body. Particularly, we propose a coarse-to-fine learning model that gradually deforms a template body toward the ground truth body. Our model utilizes the information of multi-view renderings and corresponding 3D vertex transformation as supervision. Such supervision will help to generate 3D bodies well aligned to all views. To accurately operate mesh deformation, a graph convolutional network structure is introduced to support the shape generation from 3D vertex representation. Additionally, a graph up-pooling operation is designed over the intermediate representations of the graph convolutional network, and thus our model can generate 3D shapes with higher resolution. Novel loss functions are employed to help optimize the whole multi-view generation model, resulting in smoother surfaces. In addition, two multi-view human body datasets are produced and contributed to the community. Extensive experiments conducted on the benchmark datasets demonstrate the efficacy of our model over the competitors.

Message flow analysis with complex causal links for distributed ROS 2 systems

Distributed robotic systems rely heavily on the publish–subscribe communication paradigm and middleware frameworks that support it, such as the Robot Operating System (ROS), to efficiently implement modular computation graphs. The ROS 2 executor, a high-level task scheduler which handles ROS 2 messages, is a performance bottleneck. We extend ros2_tracing, a framework with instrumentation and tools for real-time tracing of ROS 2, with the analysis and visualization of the flow of messages across distributed ROS 2 systems. Our method detects one-to-many and many-to-many causal links between input and output messages, including indirect causal links through simple user-level annotations. We validate our method on both synthetic and real robotic systems, and demonstrate its low runtime overhead. Moreover, the underlying intermediate execution representation database can be further leveraged to extract additional metrics and high-level results. This can provide valuable timing and scheduling information to further study and improve the ROS 2 executor as well as optimize any ROS 2 system. The source code is available at: github.com/christophebedard/ros2-message-flow-analysis.

Integrated extreme gradient boost with c4.5 classifier for high level synthesis in very large scale integration circuits

High-level synthesis (HLS) is utilized for high-performance and energy-efficient heterogeneous systems designing. HLS is assist in field-programmable gate array circuits designing where hardware implementations are refined and replaced in target device. However, the power-process-voltage-temperature-delay (PPVTD) variation in VLSI circuits undergoes many problems and reduced the performance. In order to address these problems, C4.5 with eXtreme Gradient Boosting Classification based High Level Synthesis (C4.5-XGBCHLS) Method is designed for afford better runtime adaptability (RA) with minimal error rate. VLSI circuits are designed using the behavioral input and results are measured at running condition. When VLSI circuit’s results get reduced, the language description of the circuit is considered as an input. Then, compilation process convert high level specification into Intermediate Representation (IR) in control/data flow graph (CDFG). CDFG computes data and control dependencies among operations. eXtreme Gradient Boosting (XGBoost) Classifier is exploited in C4.5-XGBCHLS method to classify the error causing functional unit (FU) with minimal error rate. XGBoost Classifier exploited C4.5 decision tree as base classifier to enhance classification of error causing FU in VLSI circuits. After that, FU gets allocated in place of error causing FU from functional library based on the design objectives and PPVTD variations. Finally, operation scheduling and binding process is executed for register transfer level (RTL) generation to form VLSI circuits with improved RA. The simulation results shows that the C4.5-XGBCHLS method enhances the performance of functional unit selection accuracy (FUSA) with minimal error rate (ER) and circuit adaptability time (CAT).

ExGen: Cross-platform, Automated Exploit Generation for Smart Contract Vulnerabilities

Smart contracts, just like other computer programs, are prone to a variety of vulnerabilities, which lead to severe consequences including massive token and coin losses. Prior works have explored automated exploit generation for vulnerable Ethereum contracts. However, the scopes of prior works are limited in both vulnerability types and contract platforms. In this paper, we propose a cross-platform framework, called <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ExGen</small> , to generate multiple transactions as exploits to given vulnerable smart contracts. <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ExGen</small> first translates either Ethereum or EOS contracts to an intermediate representation (IR). Then, <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ExGen</small> generates symbolic attack contracts with transactions in a partial order and then symbolically executes the attack contracts together with the target to find and solve all the constraints. Lastly, <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ExGen</small> concretizes all the symbols, generates attack contracts with multiple transactions, and verifies the generated contracts’ exploitability on a private chain with values crawled from the public chain. We implemented a prototype of <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ExGen</small> and evaluated it on Ethereum and EOS benchmarks. <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ExGen</small> successfully exploits 1,258/1,399 (89.9%) Ethereum and 126/130 (96.9%) EOS vulnerabilities. <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ExGen</small> is also able to exploit zero-day vulnerabilities on EOS.

Canal: A Flexible Interconnect Generator for Coarse-Grained Reconfigurable Arrays

The architecture of a coarse-grained reconfigurable array (CGRA) interconnect has a significant effect on not only the flexibility of the resulting accelerator, but also its power, performance, and area. Design decisions that have complex trade-offs need to be explored to maintain efficiency and performance across a variety of evolving applications. This paper presents Canal, a Python-embedded domain-specific language (eDSL) and compiler for specifying and generating reconfigurable interconnects for CGRAs. Canal uses a graph-based intermediate representation (IR) that allows for easy hardware generation and tight integration with place and route tools. We evaluate Canal by constructing both a fully static interconnect and a hybrid interconnect with ready-valid signaling, and by conducting design space exploration of the interconnect architecture by modifying the switch box topology, the number of routing tracks, and the interconnect tile connections. Through the use of a graph-based IR for CGRA interconnects, the eDSL, and the interconnect generation system, Canal enables fast design space exploration and creation of CGRA interconnects.

MemBridge: Video-Language Pre-training with Memory-Augmented Inter-Modality Bridge.

Video-language pre-training has attracted considerable attention recently for its promising performance on various downstream tasks. Most existing methods utilize the modality-specific or modality-joint representation architectures for the cross-modality pre-training. Different from previous methods, this paper presents a novel architecture named Memory-augmented Inter-Modality Bridge (MemBridge), which uses the learnable intermediate modality representations as the bridge for the interaction between videos and language. Specifically, in the transformer-based cross-modality encoder, we introduce the learnable bridge tokens as the interaction approach, which means the video and language tokens can only perceive information from bridge tokens and themselves. Moreover, a memory bank is proposed to store abundant modality interaction information for adaptively generating bridge tokens according to different cases, enhancing the capacity and robustness of the inter-modality bridge. Through pre-training, MemBridge explicitly models the representations for more sufficient inter-modality interaction. Comprehensive experiments show that our approach achieves competitive performance with previous methods on various downstream tasks including video-text retrieval, video captioning, and video question answering on multiple datasets, demonstrating the effectiveness of the proposed method. The code has been available at https://github.com/jahhaoyang/MemBridge.

Dual Branch PnP Based Network for Monocular 6D Pose Estimation

Monocular 6D pose estimation is a functional task in the field of computer vision and robotics. In recent years, 2D-3D correspondence-based methods have achieved improved performance in multiview and depth data-based scenes. However, for monocular 6D pose estimation, these methods are affected by the prediction results of the 2D-3D correspondences and the robustness of the perspective-n-point (PnP) algorithm. There is still a difference in the distance from the expected estimation effect. To obtain a more effective feature representation result, edge enhancement is proposed to increase the shape information of the object by analyzing the influence of inaccurate 2D-3D matching on 6D pose regression and comparing the effectiveness of the intermediate representation. Furthermore, although the transformation matrix is composed of rotation and translation matrices from 3D model points to 2D pixel points, the two variables are essentially different and the same network cannot be used for both variables in the regression process. Therefore, to improve the effectiveness of the PnP algorithm, this paper designs a dual-branch PnP network to predict rotation and translation information. Finally, the proposed method is verified on the public LM, LM-O and YCB-Video datasets. The ADD(S) values of the proposed method are 94.2 and 62.84 on the LM and LM-O datasets, respectively. The AUC of ADD(-S) value on YCB-Video is 81.1. These experimental results show that the performance of the proposed method is superior to that of similar methods.

Programming Language Conversion Using NLP

Natural language processing strives to build machines that understand and respond to text or voice data—and respond with text or speech of their own—in much the same way humans do. NLP models language computationally and deals with linguistic features of computation. Once a computer learns to do mathematical calculations it can perform many complex and big calculations much faster than humans. Similarly once computer starts to understand the human languages it can process all aspects of that language much faster than humans also opening a large number of possibility. So it cuts down on employment as one computer is capable of giving an output 10 times faster than a human can. So it benefits the employer not only financially but also by giving extremely accurate and faster outcome. Here we lay out an overall architecture to explain the overall processing. So now we take a look at the two general classes of systems they are special-purpose system and general-purpose system, explaining how they differ and their relative advantages and disadvantages. After that we point at the few remaining problems that require additional research. Finally, we conclude by discussing when natural language processing technology can be practically used at various levels . We also discuss about when it will become commercially practical, and what will be the cost to practically use this technology. The techniques specifically developed for analysing and understanding the inner-workings and representations acquired by neural models of language is EMNLP 2018 BlackboxNLP. The approach includes: investigating the impact on the performance of neural network on systematic manipulation of input and also testing whether the interpretable knowledge can be decoded from intermediate representations to propose modifications to make the knowledge state or generated output more and also to examine the performance of networks on simplified or formal languages. In the following report we aim to convert a program of a given language to an equivalent program of another language. For that we have taken help of NLP that is Natural Language Processing. By using Natural Language Tool Kit, we have successfully identified the variables, datatypes, operators, keywords, indentations. We have also discussed various aspects and domains of NLP and some real-world applications of it.

Noise-Separated Adaptive Feature Distillation for Robust Speech Recognition

This letter makes an improvement on feature-based knowledge distillation for robust speech recognition. The use of distillation techniques in speech recognition has been demonstrated to improve the robustness of the system. In this letter, we propose a noise-separated adaptive feature distillation method, including an adaptive distillation position selection strategy and a noise separation mechanism, assuming that there is a common network structure between the student and teacher. The proposed method has two improvements. First, distillation positions can be adaptively selected in each iteration by comparing loss values computed on intermediate representations of the student and the teacher, increasing the flexibility of knowledge transfer during distillation. Second, a noise separation module is proposed to constrain noise information elimination by explicitly separating the speech information and the noise information in noisy speech, which reduces the interference of noise information during distillation. Therefore, a better recognition performance is demonstrated with the proposed method compared to the standardized feature-based knowledge distillation method.

ViT VO - A Visual Odometry technique Using CNN-Transformer Hybrid Architecture

Localization is one of the main tasks involved in the operation of autonomous agents (e.g., vehicle, robot etc.). It allows them to be able to track their paths and properly detect and avoid obstacles. Visual Odometry (VO) is one of the techniques used for agent localization. VO involves estimating the motion of an agent using the images taken by cameras attached to it. Conventional VO algorithms require specific workarounds for challenges posed by the working environment and the captured sensor data. On the other hand, Deep Learning approaches have shown tremendous efficiency and accuracy in tasks that require high degree of adaptability and scalability. In this work, a novel deep learning model is proposed to perform VO tasks for space robotic applications. The model consists of an optical flow estimation module which abstracts away scene-specific details from the input video sequence and produces an intermediate representation. The CNN module which follows next learn relative poses from the optical flow estimates. The final module is a state-of-the-art Vision Transformer, which learn absolute pose from the relative pose learnt by the CNN module. The model is trained on the KITTI dataset and has obtained a promising accuracy of approximately 2%. It has outperformed the baseline model, MagicVO, in a few sequences in the dataset.

Статический анализ на основе обобщённого абстрактного синтаксического дерева

The paper describes a unified representation for an abstract syntax tree (AST) suitable for static analysis of several programming languages. The proposed analysis scheme consists of saving an intermediate representation in the form of a unified AST from compilers of the corresponding languages and subsequent analysis of the saved trees. We have implemented this described representation for Java, Kotlin and Python. The unified AST analyzer has 27 checkers. In the paper we present structure and entities of our unified AST, provide more details regarding language specifics that have to be reflected in the UAST representation. We give extensive experimental results that show UAST generation and analysis speed, analysis quality, and comparison with the old scheme of analyzing compiler ASTs where applicable. As a result, we see that we observe some degradation of analysis speed, but we pay it for the separation of AST construction and checkers’ implementation. This separation allows easier support of many languages in the analyzer, where one can just generate UAST and support the required checker once within the UAST infrastructure instead of implementing a checker once per language.

Generalizable Pancreas Segmentation Modeling in CT Imaging via Meta-Learning and Latent-Space Feature Flow Generation.

Accurate pancreas segmentation is highly crucial for diagnosing and treating pancreatic diseases. Although CNN has demonstrated promising outcomes, the performance on unseen data can be significantly compromised by the wide appearance-style variations induced by different imaging factors. Thus, we propose a generalizable pancreas segmentation model based on a meta-learning strategy and latent-space feature flow generation method. Our approach enhances the generalizability by systematically reducing the interference from the cluttered background and appearance-style discrepancies through a coarse-to-fine workflow. Specifically, the integrity-preserving coarse segmentation module is designed to adaptively balance the pancreas coverage and segmentation accuracy with the meta-learning strategy for filtering out background clutter. It also enhances the generalization of the coarse model to reasonably-accurate ROIs thereby promoting the stability of fine segmentation. Subsequently, the appearance-style feature flow generation method is developed to generate a series of progressively-varying style-related intermediate representations between two latent spaces. This feature flow effectively models the distribution variations caused by appearance-style discrepancies, and thus enhances the adaptability of the fine model. Our method achieves superior performance on three pancreas datasets and outperforms state-of-the-art generalization methods. Besides, it can be easily integrated into other workflows, leading to a potential paradigm for enhancing generalization performance.

IsQ: An Integrated Software Stack for Quantum Programming

We introduce isQ, a new software stack for quantum programming in an imperative programming language, also named isQ. The aim of isQ is to make programmers write quantum programs as conveniently as possible. In particular: <list list-type="ordered" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <list-item><label>1)</label> The isQ language and its compiler contain many useful features, including but not limited to: classical control flow, such as recursion; decomposition of self-defined unitary gates; and oracle programming and its circuit realization. </list-item> <list-item><label>2)</label> To make it flexible, an isQ program can be compiled into several different kinds of intermediate representation and assemblies, including QIR, eQASM, OpenQASM 3.0, and QCIS (specially tailored for the superconducting quantum hardware at USTC). </list-item> <list-item><label>3)</label> Besides interfacing isQ with real superconducting hardware, a QIR simulator is also developed for the demonstration and testing of isQ programs. </list-item> </list> <p xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">isQ software stack encompasses abundant compiler optimizations of high-level quantum programs. To realize it, a distinct MLIR dialect name isQ-IR is proposed.

Learning Domain and Pose Invariance for Thermal-to-Visible Face Recognition

Interest in thermal to visible face recognition has grown significantly over the last decade due to advancements in thermal infrared cameras and analytics beyond the visible spectrum. Despite large discrepancies between thermal and visible spectra, existing approaches bridge domain gaps by either synthesizing visible faces from thermal faces or by learning the cross-spectrum image representations. These approaches typically work well with frontal facial imagery collected at varying ranges and expressions, but exhibit significantly reduced performance when matching thermal faces with varying poses to frontal visible faces. We propose a novel Domain and Pose Invariant Framework that simultaneously learns domain and pose invariant representations. Our proposed framework is composed of modified networks for extracting the most correlated intermediate representations from off-pose thermal and frontal visible face imagery, a sub-network to jointly bridge domain and pose gaps, and a joint-loss function comprised of cross-spectrum and pose-correction losses. We demonstrate efficacy and advantages of the proposed method by evaluating on three thermal-visible datasets: ARL Visible-to-Thermal Face, ARL Multimodal Face, and Tufts Face. Although DPIF focuses on learning to match off-pose thermal to frontal visible faces, we also show that DPIF enhances performance when matching frontal thermal face images to frontal visible face images.

Speech translation for Unwritten language using intermediate representation: Experiment for Viet-Muong language pair

The paper studies an automatic translation method that translates from the text of a language (L1) to the speech of an unwritten language (L2). Normally the written text is used as the bridge to connect a translation module that translates from the text of L1 to the text of L2 and a synthesis module that generates the speech of L2 from the text. In the case of unwritten language, an intermediate representation has to be used instead of the writing form of L2. This paper proposes the use of phoneme representation because of the intimate relationship between phonemes and speech in one language. The proposed method was applied to the Viet-Muong language pair. The Vietnamese text needs to be translated into Muong language in two dialects, Muong Bi - Hoa Binh and Muong Tan Son - Phu Tho, both unwritten. The paper also proposes a phoneme set for each Muong language and applies them to the problem. The evaluation results showed that the translation quality was relatively high in both dialects (for Muong Bi, the fluency score was 4.63/5.0, and the adequacy score was 4.56/5.0). The synthesized speaking quality in both dialects is acceptable (for Muong Bi, the MOS score was 4.47/5.0, and the comprehension score was 93.55%). The results also show that the applicability of the proposed system to other unwritten languages ​​is promising.