Recognizing Functions in Binaries with Neural Networks

By Eui Chul Richard Shin, Dawn Song, Reza Moazzezi. 2015

317 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Neural Cleanse: Identifying and Mitigating Backdoor Attacks in Neural Networks (2019, methodology): “In the security space, DNNs are used for everything from malware classification [1], [2], to binary reverseengineering [3], [4] and network intrusion detection [5].”
Practical Black-Box Attacks against Machine Learning (2016, background): “4]. Thus, deep learning has been broadly adapted to automotive systems [29], ﬁnance [22], healthcare [1], computer vision [23], speech recognition [18], natural language processing [32], and security [11, 36]. It has been found that a wide variety of machine learning algorithms are vulnerable to int”
discovRE: Efficient Cross-Architecture Identification of Bugs in Binary Code (2016, methodology): “To further improve these results, one can leverage the recent and advanced academic methods to recognize functions in binary code such as [16, 55].”
Practical Black-Box Attacks against Deep Learning Systems using Adversarial Examples (2016, background): “Thus, deep learning has been broadly adapted to automotive systems [29], finance [22], healthcare [1], computer vision [23], speech recognition [18], natural language processing [32], and security [11, 36].”
Statistical Deobfuscation of Android Applications (2016, methodology): “[34] uses neural networks and a large training set to identify libraries in binaries.”
SySeVR: A Framework for Using Deep Learning to Detect Software Vulnerabilities (2018, methodology): “CNN has been used for software defect prediction [16] and locating buggy source code [51]; DBN has been used for software defect prediction [19], [20]; RNN has been used for vulnerability detection [11], [15], [50], software traceability [12], code clone detection [13], and recognizing functions in ”
Neural Nets Can Learn Function Type Signatures From Binaries (2017, methodology): “Previous approaches range from traditional machine learning techniques [7] to neural networks [43] to applying function interface verification [35].”
LEMNA: Explaining Deep Learning based Security Applications (2018, methodology): “Recently, researchers show that well-trained RNN can help handle critical reverse-engineering steps such as detecting the function start [52], which can significantly save human efforts.”
Learning to Fuzz from Symbolic Execution with Application to Smart Contracts (2019, background): “[55] learn to recognize functions in stripped binaries.”
Automated Crowdturfing Attacks and Defenses in Online Review Systems (2017, background): “Work in this category mainly includes proposals to improve network intrusion detection [8, 25], malware classification [62, 66, 81], and passwordbased authentication systems [39].”
Neural Network-based Graph Embedding for Cross-Platform Binary Code Similarity Detection (2017, result): “In recent years, deep learning [28] has been applied to many application domains, including binary analysis [42], and has shown stronger results than other approaches.”
DeepSim: deep learning code functional similarity (2018, background): “Recently, a few efforts [8, 48, 58] have also been made to tackle program analysis problems with DNN.”
Unsolved Problems in ML Safety (2021, background): “Massive unsupervised ML methods could also model binaries and learn to detect malicious obfuscated payloads [168, 161, 134, 71].”
Scalable Graph-based Bug Search for Firmware Images (2016, methodology): “Although IDA pro [28] provides us reasonable accuracy in our evaluation, we can rely on more advanced techniques to further improve its accuracy such as [51].”
Using Non-invertible Data Transformations to Build Adversary-Resistant Deep Neural Networks (2016, background): “In cybersecurity, security companies have demonstrated that deep learning could offer a far better way to detect all types of malware [27], [4], [38] and recognize such functions in binary code [32].”
Using Non-invertible Data Transformations to Build Adversarial-Robust Neural Networks (2016, background): “In cybersecurity, security companies have demonstrated that deep learning could offer a far better way to detect all types of malware [27], [4], [38] and recognize such functions in binary code [32].”