I work on machine learning security, malware detection, adversarial robustness, and computer vision robustness. My research focuses on building and evaluating systems under realistic adversarial and corrupted conditions.
ICPR 2026. To be presented in August 2026.
ImageNet-LC is an object-centric robustness benchmark for ImageNet. Instead of applying corruptions globally, it localizes foreground object regions and applies corruptions only inside those regions of interest.
- Official code: https://github.com/muskanny/ImageNet-LC
- Repository pointer: https://github.com/cssanchit/ImageNet-LC
- Dataset: https://www.kaggle.com/datasets/sanchitgupta10/imagenet-pc
ICISS 2024.
REMEDII studies robust malware detection under GAN-based adversarial malware examples and iterative adversarial training.
- Paper: https://dl.acm.org/doi/10.1007/978-3-031-80020-7_14
- Code: https://github.com/cssanchit/REMEDII-Malware-Detection
A reusable public repository of CNN-based face-recognition transfer attacks with paper references, adaptation notes, contributor credits, and attack-specific documentation. It is designed to help researchers and students inspect, reuse, and extend transfer attacks in the face-verification setting.
- ImageNet-LC: Assessing Robustness under Localized Corruptions. ICPR 2026.
- REMEDII: Robust Malware Detection with Iterative and Intelligent Adversarial Training. ICISS 2024. https://dl.acm.org/doi/10.1007/978-3-031-80020-7_14
- Characterizing and Classifying Android Malware: A High-Level Feature Approach. 2023 International Conference on Quantum Technologies, Communications, Computing, Hardware and Embedded Systems Security (iQ-CCHESS). https://ieeexplore.ieee.org/document/10391357
- LFSR Next Bit Prediction through Deep Learning. Journal of Informatics Electrical and Electronics Engineering, 2021. https://doi.org/10.54060/JIEEE/002.02.022
- Malware Detection in PDF and Office Documents: A Survey. Information Security Journal: A Global Perspective, 2020. https://doi.org/10.1080/19393555.2020.1723747
- Malware Characterization Using Windows API Call Sequences. Security, Privacy, and Applied Cryptography Engineering (SPACE 2016); Journal of Cyber Security and Mobility, 2018. https://link.springer.com/chapter/10.1007/978-3-319-49445-6_15
- Impact of Machine Learning Algorithms on Analysis of Stream Ciphers. International Conference on Methods and Models in Computer Science (ICM2CS), 2009. https://ieeexplore.ieee.org/document/5397953
- Adversarial machine learning
- Malware detection and malware characterization
- Robustness benchmarking
- Computer vision robustness
- Security applications of machine learning
- M.Tech in Computer Science and Engineering, IIT Delhi.
- B.Tech in Computer Science and Engineering, NIT Hamirpur.
- GitHub: https://github.com/cssanchit
- Kaggle dataset: https://www.kaggle.com/datasets/sanchitgupta10/imagenet-pc