An Xception Convolutional Neural Network for Malware Classification with Transfer Learning
The DOI matches its Crossref record; 11 of the 11 references it cites were found in a registry or an index.
Abstract
In this work, we applied a deep Convolutional Neural Network (CNN) with Xception model to perform malware image classification. The Xception model is a recently developed special CNN architecture that is more powerful with less over- fitting problems than the current popular CNN models such as VGG16. However only a few use cases of the Xception model can be found in literature, and it has never been used to solve the malware classification problem. The performance of our approach was compared with other methods including KNN, SVM, VGG16 etc. The experiments on two datasets (Malimg and Microsoft Malware Dataset) demonstrated that the Xception model can achieve the highest training accuracy than all other approaches including the champion approach, and highest validation accuracy than all other approaches including VGG16 model which are using image-based malware classification (except the champion solution as this information was not provided). Additionally, we proposed a novel ensemble model to combine the predictions from .bytes files and .asm files, showing that a lower logloss can be achieved. Although the champion on the Microsoft Malware Dataset achieved a bit lower logloss, our approach does not require any features engineering, making it more effective to adapt to any future evolution in malware, and very much less time consuming than the champion's solution.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Conversely, the suggested Xception transformation approach achieves extra learning efficiency and better performance but does not minimize the number of parameters [126, 127].”
From Review of deep learning: concepts, CNN architectures, challenges, applications, future directions (2021) · cites this paper as [126]
Citations
Cites
Xception: Deep Learning with Depthwise Separable Convolutions
Chollet, F. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.195 · Cited by 20,380 · cited 1×
“The Xception model [14] was proposed by Google, Inc.”
MethodXception: Deep Learning with Depthwise Separable Convolutions
Chollet, F. (2016) · arXiv · doi:10.48550/arxiv.1610.02357 · Cited by 538 · cited 1×
“The Xception model [14] was proposed by Google, Inc.”
MethodMalware Classification with Deep Convolutional Neural Networks
Kalash, M., Rochan, M., Mohammed, N., Bruce, N. D. B., Wang, Y., & Iqbal, F. (2018) · 2018 9th IFIP International Conference on New Technologies, Mobility and Security (NTMS) · doi:10.1109/ntms.2018.8328749 · Cited by 408 · cited 2×
“The experiments results demonstrate that the validation accuracy of the Xception model on the Malimg Dataset [8] and the Microsoft malware datasets [9] is higher (99.”
MethodSelf-citationMicrosoft Malware Classification ChallengePreprint
Ronen, R., Radu, M., Feuerstein, C., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1802.10135 · Cited by 287 · cited 1×
“Microsoft Malware Dataset The Microsoft Malware Dataset [15] was released by the Kaaggle Microsoft Malware Classification Competition.”
MethodMicrosoft Malware Classification Challenge
Ronen, R., Radu, M., Feuerstein, C., Yom‐Tov, E., & Ahmadi, M. (2018) · arXiv (Cornell University) · doi:10.13140/rg.2.2.34695.91045 · Cited by 286
Malware images: visualization and automatic classification
Nataraj (2011) · (No Title) · Cited by 215
Visual Reverse Engineering of Binary and Data Files
Conti, G., Dean, E., Sinda, M., & Sangster, B. (2008) · Lecture notes in computer science · doi:10.1007/978-3-540-85933-8_1 · Cited by 93 · cited 1×
“[17] visualized the raw binary data of primitive binary fragments (such as text, C++ data structures, image data, and video data) into the images.”
BackgroundMalware Images Classification Using Convolutional Neural Network
Kabanga, E. K., & Kim, C. (2017) · Journal of Computer and Communications · doi:10.4236/jcc.2018.61016 · Cited by 70
Classification of malware families based on N-grams sequential pattern features
-, C. L., & Sornil, O. (2013) · IEEE Conference on Industrial Electronics and Applications (ICIEA) · doi:10.1109/iciea.2013.6566472 · Cited by 61 · cited 2×
“Nowadays, most antivirus vendors have used machine learning to solve this problem and applied these techniques throughout the industry, for example using statistical characteristics like API calls [4] or N-grams [5].”
MethodPolymorphic malware detection using sequence classification methods and ensembles
Drew, J. M., Hahsler, M., & Moore, T. (2017) · EURASIP Journal on Information Security · doi:10.1186/s13635-017-0055-6 · Cited by 59 · cited 1×
“Strand Gene Sequence [18] .”
Research works printed but not linked: 8
- lin, 2013, Network in Network
- makandar, 2015, Malware Image Analysis and Classification using Support Vector Machine, International Journal of Advanced Trends in Computer Science and Engineering, 4, 1
- team, 2015, Microsoft Malware Winners’ Interview 1st place ‘NO to overfitting! - Confusion matrix’ | No Free Hunch
- li, 2017, CS231 Lecture 09? CNN Architectures, 1
- 2017, How Traditional Antivirus Works
- 2018, Documentation for individual models
- 2017, Kaspersky Lab detects 360 000 new malicious files daily – up 11 5% from 2016
- robert moir, 2003, Defining Malware
Cited by
Items 101–129 of 129
Contextualized Vector Embeddings for Malware Detection
Pandya, V. (2022) · San José State University ScholarWorks (San Jose State University) · doi:10.31979/etd.rjra-9c8m · Cited by 2
“Even though CNN’s are heavily used for image analysis, they can be used for malware classification as seen in [28].”
MethodEnhanced Balanced-Generative Adversarial Networks to Predict Pedestrian Injury TypesPreprint
Somvanshi, S., Antariksa, G., & Das, S. (2024) · Research Square · doi:10.21203/rs.3.rs-4530893/v1 · Cited by 2
Najaflou, S., & Lesani, F. S. (2023) · doi:10.1109/ipria59240.2023.10147183 · Cited by 1
“Since 2017, the Xception is used to solve various problems successfully [7,8].”
BackgroundMethodBrain Cysts and Cystic Tumors Classification Based on New Deep Learning Hybrid Structure
Coşkun, O., Özaltın, Ö., Yonar, A., & Yeniay, Ö. (2025) · Iranian Journal of Science · doi:10.1007/s40995-025-01824-y · Cited by 1
“It possesses 55.9 million parameters and 164 depths with 36 convolutional layers (Lo et al. 2019; Sharma and Kumar 2022).”
InfluentialMulti-head Attention-Aware Deep Ensemble Learning for IoT Malware Detection
Virmani, S., Bhushan, K., & Rehman, M. Z. U. (2025) · Lecture notes in networks and systems · doi:10.1007/978-981-96-2299-3_24 · Cited by 1
Diffusion-Aware Deepfake Detection via Kernel PCA on Noise Residuals for Semantic Communication
Chen, B. (2025) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2025-spring65109.2025.11174885 · Cited by 1
“In 2019, researchers introduced the XceptionNet architecture [21], which improved the detection of GAN-generated images.”
AMALGAN: Image‐Based Android Malware Classification Using Generative Adversarial Network
Qaisar, Z. H., Faheem, M., & Ashraf, M. W. (2025) · The Journal of Engineering · doi:10.1049/tje2.70146 · Cited by 1
Efficacy of Transfer Learning Over Semantic Segmentation
Shyam, A., & Selvam, S. (2022) · Communications in computer and information science · doi:10.1007/978-3-030-96040-7_52 · Cited by 1
Transfer Learning-Based Malware Classification
Chakraborty, A., & Kumar, S. (2023) · Lecture notes in networks and systems · doi:10.1007/978-981-19-6525-8_3 · Cited by 2
Potato Blight Classification Android Application using Deep learning
Chapa, M. K., Mehta, N. K., Manikanta, P., Chaitanya, U., & Sakunthala, Y. (2023) · International Journal of Advanced Research in Science Communication and Technology · doi:10.48175/ijarsct-9598 · Cited by 1
Ensemble Malware Classifier Considering PE Section Information
Takeuchi, R., Mitsuhashi, R., Nishigaki, M., & Ohki, T. (2023) · IEICE Transactions on Fundamentals of Electronics Communications and Computer Sciences · doi:10.1587/transfun.2023cip0024 · Cited by 2
Prawiranata, F. P. S., & Hadiprakoso, R. B. (2023) · IEEE International Conference on Cryptography, Informatics, and Cybersecurity (ICoCICs) · doi:10.1109/icocics58778.2023.10277454 · Cited by 1
“Several studies have demonstrated the success of transfer learning in malware classification using models such as InceptionV3, Xception, and EfficientNet [8], [10], [15].”
BackgroundMethodInfluentialOptimized Insect Classification on Farms using Tuned Convolutional Neural Networks
Moonis, A., & Singh, A. P. (2024) · International Conference on Computing, Communication and Networking Technologies (ICCCNT) · doi:10.1109/icccnt61001.2024.10726116 · Cited by 1
“Lo et al. explored the use of Xception model in solving malware classification [16] and demonstrated superiority of Xception over other methods like SVM, KNN etc.”
MethodClassification of Chili Varieties Using Convolutional Neural Network Model
Rangkuti, A. H., Wardana, L. M. P., Quincy, A., Pohan, H. I., & Maheswari, F. S. (2024) · International Conference on Cybernetics and Intelligent System (ICORIS) · doi:10.1109/icoris63540.2024.10903854 · Cited by 1
Malware Classification Based on Hidden Markov Model and Word2Vec Features
Kale, A. S. (2020) · San José State University ScholarWorks (San Jose State University) · doi:10.31979/etd.edkg-dtq8
Windows Malware Hunting with InceptionResNetv2 Assisted Malware Visualization Approach
Sharma, O., Sharma, A., & Kalia, A. (2023) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-981-99-0609-3_12
Performance Evaluation of Different Deep Learning Models for Breast Cancer Detection in Mammograms
Das, J. K., Pramanik, S., & Bhattacharjee, D. (2023) · Algorithms for intelligent systems · doi:10.1007/978-981-99-1509-5_30
Predicting Method for Security Situation of Edge Devices in Power Internet of Things
Zhang, D., Han, X., Wang, Y., Sun, X., Han, J., & Zhu, X. (2023) · IEEE International Conference on Computer and Communications (ICCC) · doi:10.1109/iccc59590.2023.10507336
“…have multiple features, high fault tolerance, self-organizing ability, and self-learning ability, and have strong nonlinear mapping and generalization ability for complex systems, Can directly construct mapping relationships for homomorphic potential values of multi-source heterogeneous data[2].”
BackgroundSelf-citationBottleneck Feature Extraction for Gene Expression Using Deep Learning
Thakur, T., Batra, I., & Malik, A. (2024) · Advances in human and social aspects of technology book series · doi:10.4018/979-8-3693-8939-3.ch011
AI-based skin cancer detection algorithms: opportunities, challenges and a way forward
Muhammad, S., Arooj, T., Adeel, H., Saleem, T., & Rao, I. R. (2025) · Asian Journal of Science Engineering and Technology (AJSET) · doi:10.47264/idea.ajset/4.1.6 · Cited by 1
Camfuse-Net: Cross Attention-Guided Multi-Scale Feature Fusion for Chest X-Ray AnalysisPreprint
Gyarteng, E. S. A., XiaoLing, Z., Ukwuoma, C. C., Agyemang, I. O., & Adjei-Mensah, I. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5338909
Malware Classification using Deep Learning on National Cyber and Crypto Agency Honeynet Data
Lesmana, D. H., & Qomariasih, N. (2025) · IEEE International Conference on Industry 4.0, Artificial Intelligence, and Communications Technology (IAICT) · doi:10.1109/iaict65714.2025.11101404
“Traditional antivirus solutions are increasingly ineffective against evolving malware threats [10].”
AL-Tahan, R. N. Y., & Talal, R. (2025) · IAES International Journal of Artificial Intelligence · doi:10.11591/ijai.v14.i4.pp2964-2978
Srinivas, P. L., Sree, T., & Surendiran, B. (2026) · Lecture notes in networks and systems · doi:10.1007/978-981-96-6329-3_12
Msipa, S. L. (2025) · Open University of Cape Town (University of Cape Town)
Şenyüz, M., Eren, F. B., & Borandağ, E. (2026) · Bilgisayar bilimleri ve mühendisliği dergisi. · doi:10.54525/bbmd.1646627
Ahmmed, F., Antik, A. R., Alaly, A., Mehnaj, S., Huda, M. S. A., & Ali, M. A. (2026) · Healthcare Technology Letters · doi:10.1049/htl2.70053
“The entry flow extracts initial low-level features, the middle flow refines them through eight repeated blocks, and the exit flow condenses the information for classification [ 60 ].”
Chandrakala, H. T., Gowda, Y. N., & Thippeswamy, G. (2026) · Cureus Journal of Computer Science. · doi:10.7759/s44389-026-00084-0
Metadata
- Authors
- Lo, W. W., Yang, X., & Wang, Y.
- Affiliations
- Macao Polytechnic University (MO)
- Venue
- 2019 10th IFIP International Conference on New Technologies, Mobility and Security (NTMS)
- Publisher
- IEEE
- Published
- 2019-06
- Pages
- 1-5
- OpenAlex
- W2960023330
- Cited by
- 122
- References
- 11
- Citations by year
- 2026: 4 · 2025: 18 · 2024: 22 · 2023: 37 · 2022: 24 · 2021: 14 · 2020: 3
- FWCI
- 7.86
- Type
- conference paper
- Language
- en
- Open access
- closed
- Licence
- Locations
- IFIP International Conference on New Technologies, Mobility and Security (NTMS) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Software › Advanced Malware Detection Techniques
- Physical Sciences › Computer Science › Hardware and Architecture › Security and Verification in Computing
- Physical Sciences › Computer Science › Information Systems › Software Engineering Research
- Keywords
- malware classification, Xception, convolutional neural networks, transfer learning, image-based malware classification, ensemble learning, Malimg dataset
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.