Investigating vulnerability of watermarking neural network
A neural network with great performance often incurs a high cost to train. The data used to train a neural network can be confidential or need additional substantial processing. Hence, a trained neural network is regarded as intellectual property. To protect a neural network from infringement of i...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project (FYP) |
| Language: | English |
| Published: |
Nanyang Technological University
2020
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/138234 |
| _version_ | 1826124388592779264 |
|---|---|
| author | Chua, Viroy Sheng Yang |
| author2 | Anupam Chattopadhyay |
| author_facet | Anupam Chattopadhyay Chua, Viroy Sheng Yang |
| author_sort | Chua, Viroy Sheng Yang |
| collection | NTU |
| description | A neural network with great performance often incurs a high cost to train. The data used to train a neural network can be confidential or need additional substantial processing. Hence, a trained neural network is regarded as intellectual property. To protect a neural network from infringement of intellectual property, the idea to watermark a neural network has been introduced. This project investigates the vulnerability of a state-of-the-art deep learning watermarking scheme. The project focus on investigating the behavior of backdoor-based watermarking scheme then proposes 2 methods to remove the watermark using the concept of transfer learning. Method 1 retrains the last convolutional layer of a model, sothe newly trained layer cannot represent the abstract features of a watermarked sample to the classifier. Method 2 involves using the basic features learn by the early convolutional layers of the watermarked model to train a model with comparable performance. The given methods show that an adversary in the same domain as the owner of the watermarked model can remove the backdoor-based watermark and invalidate any potential claim on the model. The investigation and methods aim to identify the vulnerability of backdoor-based watermark so a countermeasure can be developed to protect neural network. |
| first_indexed | 2024-10-01T06:19:49Z |
| format | Final Year Project (FYP) |
| id | ntu-10356/138234 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T06:19:49Z |
| publishDate | 2020 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1382342020-04-29T06:46:18Z Investigating vulnerability of watermarking neural network Chua, Viroy Sheng Yang Anupam Chattopadhyay School of Computer Science and Engineering anupam@ntu.edu.sg Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence A neural network with great performance often incurs a high cost to train. The data used to train a neural network can be confidential or need additional substantial processing. Hence, a trained neural network is regarded as intellectual property. To protect a neural network from infringement of intellectual property, the idea to watermark a neural network has been introduced. This project investigates the vulnerability of a state-of-the-art deep learning watermarking scheme. The project focus on investigating the behavior of backdoor-based watermarking scheme then proposes 2 methods to remove the watermark using the concept of transfer learning. Method 1 retrains the last convolutional layer of a model, sothe newly trained layer cannot represent the abstract features of a watermarked sample to the classifier. Method 2 involves using the basic features learn by the early convolutional layers of the watermarked model to train a model with comparable performance. The given methods show that an adversary in the same domain as the owner of the watermarked model can remove the backdoor-based watermark and invalidate any potential claim on the model. The investigation and methods aim to identify the vulnerability of backdoor-based watermark so a countermeasure can be developed to protect neural network. Bachelor of Engineering (Computer Science) 2020-04-29T06:46:18Z 2020-04-29T06:46:18Z 2020 Final Year Project (FYP) https://hdl.handle.net/10356/138234 en application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence Chua, Viroy Sheng Yang Investigating vulnerability of watermarking neural network |
| title | Investigating vulnerability of watermarking neural network |
| title_full | Investigating vulnerability of watermarking neural network |
| title_fullStr | Investigating vulnerability of watermarking neural network |
| title_full_unstemmed | Investigating vulnerability of watermarking neural network |
| title_short | Investigating vulnerability of watermarking neural network |
| title_sort | investigating vulnerability of watermarking neural network |
| topic | Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence |
| url | https://hdl.handle.net/10356/138234 |
| work_keys_str_mv | AT chuaviroyshengyang investigatingvulnerabilityofwatermarkingneuralnetwork |