Abstract:The continuous emergence of malware has threatened to the Android platform and user privacy. With the evolution of the Android system and malware, it is challenging to design a method that can accurately identify the categories of sophisticated malware, including known and unknown families, as well as their obfuscated variants, given that they may be newly emerging and lack available detection knowledge. Although some methods try to use anomaly detection and zero-shot technology to identify unseen applications, they are limited to binary classification or lack the robustness, stability, universality, and interpretability in multi-class identification. To this end, we first propose a generic meta-features mining algorithm, which can discover the potential relationships between samples belonging to the same category. Then we present metaNet, a novel method leveraging meta-features to identify sophisticated Android malware. Specifically, metaNet is mainly powered by four components: (i) mExtractor is a feature collector to obtain the static and dynamic features. (ii) mProcessor is taking unique meta-features of each category from extracted features. (iii) mLearner is a machine learning suite that leverages features and meta-features to design and train a classifier called HSU-Net. (iv) mEnforcer is a flexible deployer that identifies categories of malware families in the real world. We implement a prototype of metaNet with 15K lines of Python code and compare it with state-of-the-art (SOTA) methods. The results show that it can not only achieve superior performance in terms of known families (99.52% of accuracy) and unknown families (99.31% of accuracy trained with 80% known families) for binary classification, but also perform well in multi-class identification, i.e., 99.05% and 93.45% of accuracy for known and unknown families, respectively. Furthermore, we deploy and evaluate metaNet in the real world. It can identify applications over an acceptable time and memory overheads, i.e., average of 11.8s and 56MB per sample with a size of 8MB. Also, the few-shot detection and feature perturbation experiments reflect its robustness and stability benefiting from meta-features. Finally, we collect the traffic of 112 decentralized applications (DApps) belonging to 16 categories, such as finance and health, and evaluated metaNet in DApp identification. The results illustrate its applicability across various tasks. That is, it can accurately classify 94.6% and 81.36% of DApp flows in all-known and 80%-known DApp scenarios, respectively, outperforming the SOTA methods.

DEMISTIFY: Identifying On-device Machine Learning Models Stealing and Reuse Vulnerabilities in Mobile Apps.

DeepMag+: Sniffing Mobile Apps in Magnetic Field Through Deep Learning.

MagThief: Stealing Private App Usage Data on Mobile Devices Via Built-in Magnetometer.

Mind Your Weight(s): A Large-scale Study on Insufficient Machine Learning Model Protection in Mobile Apps

Robustness of on-Device Models: Adversarial Attack to Deep Learning Models on Android Apps

AirBag: Boosting Smartphone Resistance to Malware Infection.

Detecting Repackaged Smartphone Applications in Third-Party Android Marketplaces

RiskRanker: scalable and accurate zero-day android malware detection.

MMGuard: Automatically Protecting On-Device Deep Learning Models in Android Apps.

Smart App Attack: Hacking Deep Learning Models in Android Apps

ModelObfuscator: Obfuscating Model Information to Protect Deployed ML-based Systems

Towards Black-box Attacks on Deep Learning Apps

A First Look at On-device Models in iOS Apps

Cheating Your Apps: Black‐box Adversarial Attacks on Deep Learning Apps

Metanet: Interpretable Unknown Mobile Malware Identification with a Novel Meta-Features Mining Algorithm

Model-Reuse Attacks on Deep Learning Systems

MalModel: Hiding Malicious Payload in Mobile Deep Learning Models with Black-box Backdoor Attack

Droid-Sec

Android HIV: A Study of Repackaging Malware for Evading Machine-Learning Detection

Poster: Sponge ML Model Attacks of Mobile Apps

Stealing Machine Learning Models via Prediction APIs