A5031166499- Software Testing and Debugging Techniques
- Software Engineering Research
- Software Reliability and Analysis Research
- Advanced Malware Detection Techniques
- Security and Verification in Computing
- Radiation Effects in Electronics
- Adversarial Robustness in Machine Learning
- Data Mining Algorithms and Applications
- Formal Methods in Verification
- Artificial Intelligence in Healthcare
- Natural Language Processing Techniques
- IoT-based Smart Home Systems
- Parallel Computing and Optimization Techniques
- Rough Sets and Fuzzy Logic
- Hand Gesture Recognition Systems
- Advanced Software Engineering Methodologies
- Smart Parking Systems Research
- Mental Health Research Topics
- Explainable Artificial Intelligence (XAI)
- Machine Learning and Data Classification
- Cryptography and Residue Arithmetic
- Stochastic Gradient Optimization Techniques
- Cloud Data Security Solutions
- Advanced Data Storage Technologies
- Digital Media Forensic Detection
Graphic Era University
2024
Chandigarh University
2024
Amity University
2024
University of Minnesota
2017-2023
Amazon (United States)
2023
University of Minnesota System
2017-2022
Microsoft Research (United Kingdom)
2022
Malaviya National Institute of Technology Jaipur
2021
Delhi Technological University
2020
Twin Cities Orthopedics
2019
Deep neural networks (DNNs) are increasingly being adopted for sensing and control functions in a variety of safety mission-critical systems such as self-driving cars, autonomous air vehicles, medical diagnostics industrial robotics. Failures can lead to loss life or property, which necessitates stringent verification validation providing high assurance. Though formal approaches investigated, testing remains the primary technique assessing dependability systems. Due nature tasks handled by...
We investigate if a mobile application running on touch-enabled device can continuously and unobtrusively authenticate identify its users based only their interactions with the user interface of application. A unique advantage that this modality provides over other implicit modalities devices is every who uses automatically enrolled into classification system via touch-based interface, thereby guaranteeing an attacker cannot avoid system. Using different types input controls available...
Merging execution paths is a powerful technique for reducing path explosion in symbolic execution. One approach, introduced and dubbed "veritesting" by Avgerinos et al., works translating abounded control flow region into single constraint. This approach convenient way to achieve merging as modification pre-existing single-path engine. Previous work evaluated this of binary code, but different design considerations apply when building tools other languages. In paper, we extend the previous Java.
We consider the problem of applying probability concepts to discover frequent itemsets in a transaction database. The paper presents probabilistic algorithm association rules. proposed outperforms apriori for larger databases without losing single rule. It involves database scan and significantly reduces number unsuccessful candidate sets generated that later fails minimum support test. uses concept recursive medians compute dispersion list each itemset. are implemented as an Inverted...
Independently developed codebases typically contain many segments of code that perform the same or closely related operations. Being able to detect these is helpful applications such as reverse engineering. In this paper, we tackle problem determining whether two binary operation by asking one segment can be substituted other. A key insight behind our approach because often have different interfaces, some glue (an adapter) will needed substitution. Here present an algorithm searches for...
We consider the problem of applying probability concepts to discover frequent itemsets in a transaction database. The paper presents probabilistic algorithm association rules. proposed outperforms priori for larger databases without losing single rule. It involves database scan and significantly reduces number unsuccessful candidate sets generated apriori that later fails minimum support test. uses concept recursive medians compute dispersion list each itemset. are implemented as an Inverted...
Object-branch coverage (OBC) is often used as a measure of the thoroughness tests suites, augmenting or substituting source-code based structural criteria such branch and modified condition/decision (MC/DC). In addition, with increasing use third-party components for which access may be unavailable, robust object-code are essential to assess how well exercised during testing. While OBC has advantage being programming language independent amenable non-intrusive measurement techniques,...
This study proposes a gesture based CAD interface that uses pose, position, velocity and direction of fingers as input data in order to draw, extrude, scale, translate rotate an object the 3D space. The system allows user generate basic geometrical primitives advanced geometries (geometries cannot be realized using CSG primitives) perform operations described above. As opposed traditional systems wherein path gestures are used carry out operations, proposed switches operate on simple binary...
Scaling symbolic execution to industrial-sized programs is an important open research problem. Veritesting a promising technique that improves scalability by combining the advantages of static with those dynamic execution. The goal veritesting reduce number paths explore in creating formulas describing regions code using disjunctive formulas. In previous work, was applied binary-level Integrating Java bytecode presents unique challenges: notably, incorporating non-local control jumps caused...
Independently developed codebases typically contain many segments of code that perform same or closely related operations (semantic clones). Finding functionally equivalent enables applications like replacing a segment by more efficient secure alternative. Such often have different interfaces, so some glue (an adapter) is needed to replace one with the other. In previous work, we presented an algorithm searches for replaceable attempting synthesize adapter between them from finite family...
Objectives: The manufactured Integrated circuits are usable only when they free from any types of faults or error. This is even more essential in complex like analog and mixed signal circuits. paper focuses on the fault classification portion using artificial neural networks. Methods/Statistical Analysis: All possible catastrophic introduced simulated an exhaustive manner Cadence Simulation package. Faults one at a time parameters were recorded. parametric variations obtained through...
Lowering costs by driving high utilization across deep learning workloads is a crucial lever for cloud providers. We present Singularity, Microsoft's globally distributed scheduling service highly-efficient and reliable execution of training inference workloads. At the heart Singularity novel, workload-aware scheduler that can transparently preempt elastically scale to drive without impacting their correctness or performance, global fleet AI accelerators (e.g., GPUs, FPGAs). All jobs in are...
Good tests are important in software development, but it can be hard to tell whether will reveal future faults that themselves unknown.Mutation analysis, which checks inserted changes a program, is strong measure of test suite adequacy, common source-or compilerlevel approaches mutation testing not applicable available only binary form.We explore analysis as an application the reassembleable disassembly approach rewriting, building tool for x86 binaries on top previously-developed Uroboros...
Existing simulations designed for cultural and interpersonal skill training rely on pre-defined responses with a menu option selection interface. Using multiple-choice interface restricting trainees' may limit the ability to apply lessons in real life situations. This systems also uses simplistic evaluation model, where selected options are marked as either correct or incorrect. model not capture sufficient information that could drive an adaptive feedback mechanism improve awareness. paper...
We address the problem of constraint encoding explosion which hinders applicability state merging in symbolic execution. Specifically, our goal is to reduce number disjunctions and if-then-else expressions introduced during merging. The main idea dynamically partition states into groups according a similar uniform structure detected their path constraints, allows efficiently encode merged memory using quantifiers. To added complexity solving quantified we propose specialized procedure that...
Object-Branch Coverage (OBC) is often used to measure effectiveness of test suites, when source code unavailable. The traditional OBC definition can be made more resilient variations in compilers and the structure generated by creating robust definitions. However finding which instructions should included each new laborious, error-prone, architecture-dependent. We automate discovery for an improved on X86 ARM architectures. discover all possible valid symbolically executing instruction...