Changes between Version 18 and Version 19 of Other/Summer/2023/SecurityAI
- Timestamp:
- Jul 25, 2023, 3:15:30 PM (16 months ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
Other/Summer/2023/SecurityAI
v18 v19 4 4 **WINLAB Summer Internship 2023** 5 5 6 **Advisors:** Tianfang Zhang, Changming Li, Hong Li6 **Advisors:** Yingying Chen, Tianfang Zhang, Changming Li, Hong Li 7 7 8 8 **Group Members:** Rut Mehta, Jacob Morin, Ethan Lung, Damon Lin 9 9 10 10 == Project Objective == 11 Artificial intelligence techniques have been widely integrated into mobile and IoT devices, enabling various functionalities based on vision (e.g., face recognition, speech recognition, and speaker identification). The extended pipeline of building deep neural networks (DNN) produces new attack surfaces, such as attacks during the data collection, model training, and model update stages. Recent research studies discovered an effective yet stealthy attack, called a backdoor attack, which trains a hidden trigger pattern into the DNNs. The backdoored DNNs will misclassify an input as an adversary-specified label if the trigger pattern appears, behaving normally otherwise, making it difficult to be detected. Backdoor attacks originate from the image domain, and recent studies have started investigating audio-domain backdoor attacks (e.g., against voice assistant systems). This project aims to study the vulnerabilities of backdoor attacks in the image and audio domains and develop techniques for attack mitigation. 11 Artificial intelligence techniques have been widely integrated into mobile and IoT devices, enabling various functionalities based on vision (e.g., face recognition, speech recognition, and speaker identification). The extended pipeline of building deep neural networks (DNN) produces new attack surfaces, such as attacks during the data collection, model training, and model update stages. Recent research studies discovered an effective yet stealthy attack, called a backdoor attack, which trains a hidden trigger pattern into the DNNs. The backdoored DNNs will misclassify an input as an adversary-specified label if the trigger pattern appears, behaving normally otherwise, making it difficult to be detected. This project focuses on improving the security behind user authentication through conventional means (e.g., passwords and facial detection) by replacing it with a biometric signature in the form of daily activities. Following this implementation, the project aims to study the vulnerabilities of backdoor attacks on such a system and develop techniques for attack mitigation. 12 13 12 14 13 15 == Week 1 == 16 **[https://docs.google.com/presentation/d/18lbTII8Xomo438rRxyvVWpK0Pbii22hXWpwpsiD9umg/edit?usp=sharing, Week 1 Presentation]** 17 14 18 **Summary** 15 19 * Established project objectives … … 19 23 20 24 21 [[Image({C52B7779-4CD1-40A3-9938-9711E34DFB86}.png )]] [[Image({C5E828CF-69BF-4DB6-9960-6A1E754AC68B}.png)]]25 [[Image({C52B7779-4CD1-40A3-9938-9711E34DFB86}.png, 200px)]] [[Image({C5E828CF-69BF-4DB6-9960-6A1E754AC68B}.png)]] 22 26 23 27 **Resources**\\ … … 26 30 27 31 [https://www.engati.com/blog/ai-for-cybersecurity#:~:text=AI%20in%20cybersecurity%20eliminates%20time,on%20more%20critical%20security%20tasks.] 28 == Week 2 32 == Week 2 == 33 **[https://docs.google.com/presentation/d/1_AcSWM-E6Nis0Qbj_dGKczmLceWf7QZA7Ns7aiDd7OY/edit?usp=sharing, Week 2 Presentation]** 34 29 35 **Summary** 30 36 * Familiarized ourselves with PyTorch 31 37 * Started researching papers about Smart User Authentication (WiFi-enabled IOT) 32 38 * Explored attack mitigation 33 [[Image({038A11F1-8AFA-4FD5-ACAA-462CE34D7381}.png)]] [[Image({30880812-ECF5-4090-B6C4-E00FFA0DF872}.png)]] 39 [[Image({30880812-ECF5-4090-B6C4-E00FFA0DF872}.png, 400px)]] 40 41 [[Image({038A11F1-8AFA-4FD5-ACAA-462CE34D7381}.png, 400px)]] 34 42 35 43 **Resources**\\ … … 39 47 [https://www.hypr.com/security-encyclopedia/iot-authentication#:~:text=IoT%20(Internet%20of%20Things)%20Authentication,%2C%20transportation%20hubs%2C%20and%20workplaces] 40 48 41 == Week 3 49 == Week 3 == 42 50 **Summary** 43 51 * Continued learning advanced PyTorch functions for IoT interference data. … … 52 60 [https://www.mdpi.com/1099-4300/23/9/1164#:~:text=The%20physical%20meaning%20of%20CSI,fading%20%5B26%2C27%5D] 53 61 54 == Week 4 62 == Week 4 == 55 63 **Summary** 56 64 * Set up Linux virtual machine through VirtualBox (Ubuntu) … … 64 72 [https://github.com/seemoo-lab/nexmon_csi#getting-started] 65 73 66 == Week 5 74 == Week 5 == 67 75 **Summary** 68 76 * Installed custom ROMs on both the Nexus 5 and Nexus 6 … … 72 80 [[Image(nexus 6 error.jpg, 420px)]] 73 81 82 == Week 6 == 74 83 84 == Week 7 == 75 85 86 == Week 8 == 76 87 88 == Week 9 == 89 90 == Week 10 ==
