zwjjiaozhu/trojai-literature

TrojAI Literature Review

The list below contains curated papers and arXiv articles that are related to Trojan attacks, backdoor attacks, and data poisoning on neural networks and machine learning systems. They are ordered "approximately" from most to least recent and articles denoted with a "*" mention the TrojAI program directly. Some of the particularly relevant papers include a summary that can be accessed by clicking the "Summary" drop down icon underneath the paper link. These articles were identified using variety of methods including:

• A flair embedding created from the arXiv CS subset; details will be provided later.
• A trained ASReview random forest model
• A curated manual literature review

1. MPAF: Model Poisoning Attacks to Federated Learning based on Fake Clients

2. PiDAn: A Coherence Optimization Approach for Backdoor Attack Detection and Mitigation in Deep Neural Networks

3. ADFL: A Poisoning Attack Defense Framework for Horizontal Federated Learning

4. Toward Realistic Backdoor Injection Attacks on DNNs using Rowhammer

5. Execute Order 66: Targeted Data Poisoning for Reinforcement Learning via Minuscule Perturbations

6. A Feature Based On-Line Detector to Remove Adversarial-Backdoors by Iterative Demarcation

7. BlindNet backdoor: Attack on deep neural network using blind watermark

8. DBIA: Data-free Backdoor Injection Attack against Transformer Networks

9. Backdoor Attack through Frequency Domain

10. NTD: Non-Transferability Enabled Backdoor Detection

11. Romoa: Robust Model Aggregation for the Resistance of Federated Learning to Model Poisoning Attacks

12. Generative strategy based backdoor attacks to 3D point clouds: Work in Progress

13. Deep Neural Backdoor in Semi-Supervised Learning: Threats and Countermeasures

14. FooBaR: Fault Fooling Backdoor Attack on Neural Network Training

15. BFClass: A Backdoor-free Text Classification Framework

16. Backdoor Attacks on Federated Learning with Lottery Ticket Hypothesis

17. Data Poisoning against Differentially-Private Learners: Attacks and Defenses

18. DOES DIFFERENTIAL PRIVACY DEFEAT DATA POISONING?

19. Check Your Other Door! Establishing Backdoor Attacks in the Frequency Domain

20. HaS-Nets: A Heal and Select Mechanism to Defend DNNs Against Backdoor Attacks for Data Collection Scenarios

21. SanitAIs: Unsupervised Data Augmentation to Sanitize Trojaned Neural Networks

22. COVID-19 Diagnosis from Chest X-Ray Images Using Convolutional Neural Networks and Effects of Data Poisoning

23. Interpretability-Guided Defense against Backdoor Attacks to Deep Neural Networks

24. Trojan Signatures in DNN Weights

25. HOW TO INJECT BACKDOORS WITH BETTER CONSISTENCY: LOGIT ANCHORING ON CLEAN DATA

26. A Synergetic Attack against Neural Network Classifiers combining Backdoor and Adversarial Examples

27. Backdoor Attack and Defense for Deep Regression

28. Use Procedural Noise to Achieve Backdoor Attack

29. Excess Capacity and Backdoor Poisoning

30. BatFL: Backdoor Detection on Federated Learning in e-Health

31. Poisonous Label Attack: Black-Box Data Poisoning Attack with Enhanced Conditional DCGAN

32. Backdoor Attacks on Network Certification via Data Poisoning

33. Identifying Physically Realizable Triggers for Backdoored Face Recognition Networks

34. Simtrojan: Stealthy Backdoor Attack

35. Back to the Drawing Board: A Critical Evaluation of Poisoning Attacks on Federated Learning

36. Quantization Backdoors to Deep Learning Models

37. Multi-Target Invisibly Trojaned Networks for Visual Recognition and Detection

38. A Countermeasure Method Using Poisonous Data Against Poisoning Attacks on IoT Machine Learning

39. FederatedReverse: A Detection and Defense Method Against Backdoor Attacks in Federated Learning

40. Accumulative Poisoning Attacks on Real-time Data

41. Inaudible Manipulation of Voice-Enabled Devices Through BackDoor Using Robust Adversarial Audio Attacks

42. Stealthy Targeted Data Poisoning Attack on Knowledge Graphs

43. BinarizedAttack: Structural Poisoning Attacks to Graph-based Anomaly Detection

44. On the Effectiveness of Poisoning against Unsupervised Domain Adaptation

45. Simple, Attack-Agnostic Defense Against Targeted Training Set Attacks Using Cosine Similarity

46. Data Poisoning Attacks Against Outcome Interpretations of Predictive Models

47. BDDR: An Effective Defense Against Textual Backdoor Attacks

48. Poisoning attacks and countermeasures in intelligent networks: status quo and prospects

49. The Devil is in the GAN: Defending Deep Generative Models Against Backdoor Attacks

50. BadEncoder: Backdoor Attacks to Pre-trainedEncoders in Self-Supervised Learning

51. BadEncoder: Backdoor Attacks to Pre-trained Encoders in Self-Supervised Learning

52. Can You Hear It? Backdoor Attacks via Ultrasonic Triggers

53. Poisoning Attacks via Generative Adversarial Text to Image Synthesis

54. Ant Hole: Data Poisoning Attack Breaking out the Boundary of Face Cluster

55. Poison Ink: Robust and Invisible Backdoor Attack

56. MT-MTD: Muti-Training based Moving Target Defense Trojaning Attack in Edged-AI network

57. Text Backdoor Detection Using An Interpretable RNN Abstract Model

58. Garbage in, Garbage out: Poisoning Attacks Disguised with Plausible Mobility in Data Aggregation

59. Classification Auto-Encoder based Detector against Diverse Data Poisoning Attacks

60. Poisoning Knowledge Graph Embeddings via Relation Inference Patterns

61. Adversarial Training Time Attack Against Discriminative and Generative Convolutional Models

62. Poisoning of Online Learning Filters: DDoS Attacks and Countermeasures

63. Rethinking Stealthiness of Backdoor Attack against NLP Models

64. Robust Learning for Data Poisoning Attacks

65. SPECTRE: Defending Against Backdoor Attacks Using Robust Statistics

66. Poisoning the Search Space in Neural Architecture Search

67. Data Poisoning Won’t Save You From Facial Recognition

68. Accumulative Poisoning Attacks on Real-time Data

69. Backdoor Attack on Machine Learning Based Android Malware Detectors

70. Understanding the Limits of Unsupervised Domain Adaptation via Data Poisoning

71. Indirect Invisible Poisoning Attacks on Domain Adaptation

72. Fight Fire with Fire: Towards Robust Recommender Systems via Adversarial Poisoning Training

73. Putting words into the system’s mouth: A targeted attack on neural machine translation using monolingual data poisoning

74. SUBNET REPLACEMENT: DEPLOYMENT-STAGE BACKDOOR ATTACK AGAINST DEEP NEURAL NETWORKS IN GRAY-BOX SETTING

75. Spinning Sequence-to-Sequence Models with Meta-Backdoors

76. Sleeper Agent: Scalable Hidden Trigger Backdoors for Neural Networks Trained from Scratch

77. Poisoning and Backdooring Contrastive Learning

78. AdvDoor: Adversarial Backdoor Attack of Deep Learning System

79. Defending against Backdoor Attacks in Natural Language Generation

80. De-Pois: An Attack-Agnostic Defense against Data Poisoning Attacks

81. Poisoning MorphNet for Clean-Label Backdoor Attack to Point Clouds

82. Provable Guarantees against Data Poisoning Using Self-Expansion and Compatibility

83. MLDS: A Dataset for Weight-Space Analysis of Neural Networks

84. Poisoning the Unlabeled Dataset of Semi-Supervised Learning

85. Regularization Can Help Mitigate Poisioning Attacks. . . With The Right Hyperparameters

86. Witches' Brew: Industrial Scale Data Poisoning via Gradient Matching

87. Towards Robustness Against Natural Language Word Substitutions

88. Concealed Data Poisoning Attacks on NLP Models

89. Covert Channel Attack to Federated Learning Systems

90. Backdoor Attacks Against Deep Learning Systems in the Physical World

91. Backdoor Attacks on Self-Supervised Learning

92. Transferable Environment Poisoning: Training-time Attack on Reinforcement Learning

93. Investigation of a differential cryptanalysis inspired approach for Trojan AI detection

94. Explanation-Guided Backdoor Poisoning Attacks Against Malware Classifiers

95. Robust Backdoor Attacks against Deep Neural Networks in Real Physical World

96. The Design and Development of a Game to Study Backdoor Poisoning Attacks: The Backdoor Game

97. A Backdoor Attack against 3D Point Cloud Classifiers

98. Explainability-based Backdoor Attacks Against Graph Neural Networks

99. DeepSweep: An Evaluation Framework for Mitigating DNN Backdoor Attacks using Data Augmentation

100. Rethinking the Backdoor Attacks' Triggers: A Frequency Perspective

101. PointBA: Towards Backdoor Attacks in 3D Point Cloud

102. Online Defense of Trojaned Models using Misattributions

103. Be Careful about Poisoned Word Embeddings: Exploring the Vulnerability of the Embedding Layers in NLP Models

104. SPECTRE: Defending Against Backdoor Attacks Using Robust Covariance Estimation

105. Black-box Detection of Backdoor Attacks with Limited Information and Data

106. TOP: Backdoor Detection in Neural Networks via Transferability of Perturbation

107. T-Miner : A Generative Approach to Defend Against Trojan Attacks on DNN-based Text Classification

108. Hidden Backdoor Attack against Semantic Segmentation Models

109. What Doesn't Kill You Makes You Robust(er): Adversarial Training against Poisons and Backdoors

110. Red Alarm for Pre-trained Models: Universal Vulnerabilities by Neuron-Level Backdoor Attacks

111. Provable Defense Against Delusive Poisoning

112. An Approach for Poisoning Attacks Against RNN-Based Cyber Anomaly Detection

113. Backdoor Scanning for Deep Neural Networks through K-Arm Optimization

114. TAD: Trigger Approximation based Black-box Trojan Detection for AI*

115. WaNet - Imperceptible Warping-based Backdoor Attack

116. Data Poisoning Attack on Deep Neural Network and Some Defense Methods

117. Baseline Pruning-Based Approach to Trojan Detection in Neural Networks*

118. Covert Model Poisoning Against Federated Learning: Algorithm Design and Optimization

119. Property Inference from Poisoning

120. TROJANZOO: Everything you ever wanted to know about neural backdoors (but were afraid to ask)

121. A Master Key Backdoor for Universal Impersonation Attack against DNN-based Face Verification

122. Detecting Universal Trigger's Adversarial Attack with Honeypot

123. ONION: A Simple and Effective Defense Against Textual Backdoor Attacks

124. Neural Attention Distillation: Erasing Backdoor Triggers from Deep Neural Networks

125. Data Poisoning Attacks to Deep Learning Based Recommender Systems

126. Backdoors hidden in facial features: a novel invisible backdoor attack against face recognition systems

127. One-to-N & N-to-One: Two Advanced Backdoor Attacks against Deep Learning Models

128. DeepPoison: Feature Transfer Based Stealthy Poisoning Attack

129. Policy Teaching via Environment Poisoning:Training-time Adversarial Attacks against Reinforcement Learning

130. Composite Backdoor Attack for Deep Neural Network by Mixing Existing Benign Features

131. SPA: Stealthy Poisoning Attack

132. Backdoor Attack with Sample-Specific Triggers

133. Explainability Matters: Backdoor Attacks on Medical Imaging

134. Escaping Backdoor Attack Detection of Deep Learning

135. Just How Toxic is Data Poisoning? A Unified Benchmark for Backdoor and Data Poisoning Attacks

136. Poisoning Attacks on Cyber Attack Detectors for Industrial Control Systems

137. Fair Detection of Poisoning Attacks in Federated Learning

138. Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxification*

139. Stealthy Poisoning Attack on Certified Robustness

140. Machine Learning with Electronic Health Records is vulnerable to Backdoor Trigger Attacks

141. Data Security for Machine Learning: Data Poisoning, Backdoor Attacks, and Defenses

142. Detection of Backdoors in Trained Classifiers Without Access to the Training Set

143. TROJANZOO: Everything you ever wanted to know about neural backdoors(but were afraid to ask)

144. HaS-Nets: A Heal and Select Mechanism to Defend DNNs Against Backdoor Attacks for Data Collection Scenarios

145. DeepSweep: An Evaluation Framework for Mitigating DNN Backdoor Attacks using Data Augmentation

146. Poison Attacks against Text Datasets with Conditional Adversarially Regularized Autoencoder

147. Strong Data Augmentation Sanitizes Poisoning and Backdoor Attacks Without an Accuracy Tradeoff

148. BaFFLe: Backdoor detection via Feedback-based Federated Learning

149. Detecting Backdoors in Neural Networks Using Novel Feature-Based Anomaly Detection

150. Mitigating Backdoor Attacks in Federated Learning

151. FaceHack: Triggering backdoored facial recognition systems using facial characteristics

152. Customizing Triggers with Concealed Data Poisoning

153. Backdoor Learning: A Survey

154. Rethinking the Trigger of Backdoor Attack

155. AEGIS: Exposing Backdoors in Robust Machine Learning Models

156. Weight Poisoning Attacks on Pre-trained Models

157. Poisoned classifiers are not only backdoored, they are fundamentally broken

158. Input-Aware Dynamic Backdoor Attack

159. Reverse Engineering Imperceptible Backdoor Attacks on Deep Neural Networks for Detection and Training Set Cleansing

160. BAAAN: Backdoor Attacks Against Autoencoder and GAN-Based Machine Learning Models

161. Don’t Trigger Me! A Triggerless Backdoor Attack Against Deep Neural Networks

162. Toward Robustness and Privacy in Federated Learning: Experimenting with Local and Central Differential Privacy

163. CLEANN: Accelerated Trojan Shield for Embedded Neural Networks

164. Witches’ Brew: Industrial Scale Data Poisoning via Gradient Matching

165. Intrinsic Certified Robustness of Bagging against Data Poisoning Attacks

166. Can Adversarial Weight Perturbations Inject Neural Backdoors?

167. Trojaning Language Models for Fun and Profit

168. Practical Detection of Trojan Neural Networks: Data-Limited and Data-Free Cases

169. Class-Oriented Poisoning Attack

170. Noise-response Analysis for Rapid Detection of Backdoors in Deep Neural Networks

171. Cassandra: Detecting Trojaned Networks from Adversarial Perturbations

172. Backdoor Learning: A Survey

173. Backdoor Attacks and Countermeasures on Deep Learning: A Comprehensive Review

174. Live Trojan Attacks on Deep Neural Networks

175. Odyssey: Creation, Analysis and Detection of Trojan Models

176. Data Poisoning Attacks Against Federated Learning Systems

177. Blind Backdoors in Deep Learning Models

178. Deep Learning Backdoors

179. Attack of the Tails: Yes, You Really Can Backdoor Federated Learning

180. Backdoor Attacks on Facial Recognition in the Physical World

181. Graph Backdoor

182. Backdoor Attacks to Graph Neural Networks

183. You Autocomplete Me: Poisoning Vulnerabilities in Neural Code Completion

184. Reflection Backdoor: A Natural Backdoor Attack on Deep Neural Networks

185. Trembling triggers: exploring the sensitivity of backdoors in DNN-based face recognition

186. Just How Toxic is Data Poisoning? A Unified Benchmark for Backdoor and Data Poisoning Attacks

187. Adversarial Machine Learning -- Industry Perspectives

188. ConFoc: Content-Focus Protection Against Trojan Attacks on Neural Networks

189. Model-Targeted Poisoning Attacks: Provable Convergence and Certified Bounds

190. Deep Partition Aggregation: Provable Defense against General Poisoning Attacks

191. The TrojAI Software Framework: An OpenSource tool for Embedding Trojans into Deep Learning Models*

192. Influence Function based Data Poisoning Attacks to Top-N Recommender Systems

193. BadNL: Backdoor Attacks Against NLP Models

     Summary

     • Introduces first example of backdoor attacks against NLP models using Char-level, Word-level, and Sentence-level triggers (these different triggers operate on the level of their descriptor)
       • Word-level trigger picks a word from the target model’s dictionary and uses it as a trigger
       • Char-level trigger uses insertion, deletion or replacement to modify a single character in a chosen word’s location (with respect to the sentence, for instance, at the start of each sentence) as the trigger.
       • Sentence-level trigger changes the grammar of the sentence and use this as the trigger
     • Authors impose an additional constraint that requires inserted triggers to not change the sentiment of text input
     • Proposed backdoor attack achieves 100% backdoor accuracy with only a drop of 0.18%, 1.26%, and 0.19% in the models utility, for the IMDB, Amazon, and Stanford Sentiment Treebank datasets

194. Neural Network Calculator for Designing Trojan Detectors*

195. Dynamic Backdoor Attacks Against Machine Learning Models

196. Vulnerabilities of Connectionist AI Applications: Evaluation and Defence

197. Backdoor Attacks on Federated Meta-Learning

198. Defending Support Vector Machines against Poisoning Attacks: the Hardness and Algorithm

199. Backdoors in Neural Models of Source Code

200. A new measure for overfitting and its implications for backdooring of deep learning

201. An Embarrassingly Simple Approach for Trojan Attack in Deep Neural Networks

202. MetaPoison: Practical General-purpose Clean-label Data Poisoning

203. Backdooring and Poisoning Neural Networks with Image-Scaling Attacks

204. Bullseye Polytope: A Scalable Clean-Label Poisoning Attack with Improved Transferability

205. On the Effectiveness of Mitigating Data Poisoning Attacks with Gradient Shaping

206. A Survey on Neural Trojans

207. STRIP: A Defence Against Trojan Attacks on Deep Neural Networks

     Summary

     • Authors introduce a run-time based trojan detection system called STRIP or STRong Intentional Pertubation which focuses on models in computer vision
     • STRIP works by intentionally perturbing incoming inputs (ie. by image blending) and then measuring entropy to determine whether the model is trojaned or not. Low entropy violates the input-dependance assumption for a clean model and thus indicates corruption
     • Authors validate STRIPs efficacy on MNIST,CIFAR10, and GTSRB acheiveing false acceptance rates of below 1%

208. TrojDRL: Trojan Attacks on Deep Reinforcement Learning Agents

209. Demon in the Variant: Statistical Analysis of DNNs for Robust Backdoor Contamination Detection

210. Regula Sub-rosa: Latent Backdoor Attacks on Deep Neural Networks

211. Februus: Input Purification Defense Against Trojan Attacks on Deep Neural Network Systems

212. TBT: Targeted Neural Network Attack with Bit Trojan

213. Bypassing Backdoor Detection Algorithms in Deep Learning

214. A backdoor attack against LSTM-based text classification systems

215. Invisible Backdoor Attacks Against Deep Neural Networks

216. Detecting AI Trojans Using Meta Neural Analysis

217. Label-Consistent Backdoor Attacks

218. Detection of Backdoors in Trained Classifiers Without Access to the Training Set

219. ABS: Scanning neural networks for back-doors by artificial brain stimulation

220. NeuronInspect: Detecting Backdoors in Neural Networks via Output Explanations

221. Universal Litmus Patterns: Revealing Backdoor Attacks in CNNs

222. Programmable Neural Network Trojan for Pre-Trained Feature Extractor

223. Demon in the Variant: Statistical Analysis of DNNs for Robust Backdoor Contamination Detection

224. TamperNN: Efficient Tampering Detection of Deployed Neural Nets

225. TABOR: A Highly Accurate Approach to Inspecting and Restoring Trojan Backdoors in AI Systems

226. Design of intentional backdoors in sequential models

227. Design and Evaluation of a Multi-Domain Trojan Detection Method on ins Neural Networks

228. Poison as a Cure: Detecting & Neutralizing Variable-Sized Backdoor Attacks in Deep Neural Networks

229. Data Poisoning Attacks on Stochastic Bandits

230. Hidden Trigger Backdoor Attacks

231. Deep Poisoning Functions: Towards Robust Privacy-safe Image Data Sharing

232. A new Backdoor Attack in CNNs by training set corruption without label poisoning

233. Deep k-NN Defense against Clean-label Data Poisoning Attacks

234. Transferable Clean-Label Poisoning Attacks on Deep Neural Nets

235. Revealing Backdoors, Post-Training, in DNN Classifiers via Novel Inference on Optimized Perturbations Inducing Group Misclassification

236. Explaining Vulnerabilities to Adversarial Machine Learning through Visual Analytics

237. Subpopulation Data Poisoning Attacks

238. TensorClog: An imperceptible poisoning attack on deep neural network applications

239. DeepInspect: A black-box trojan detection and mitigation framework for deep neural networks

240. Resilience of Pruned Neural Network Against Poisoning Attack

241. Spectrum Data Poisoning with Adversarial Deep Learning

242. Neural cleanse: Identifying and mitigating backdoor attacks in neural networks

243. SentiNet: Detecting Localized Universal Attacks Against Deep Learning Systems

     Summary

     • Authors develop SentiNet detection framework for locating universal attacks on neural networks
     • SentiNet is ambivalent to the attack vectors and uses model visualization / object detection techniques to extract potential attacks regions from the models input images. The potential attacks regions are identified as being the parts that influence the prediction the most. After extraction, SentiNet applies these regions to benign inputs and uses the original model to analyze the output
     • Authors stress test the SentiNet framework on three different types of attacks— data poisoning attacks, Trojan attacks, and adversarial patches. They are able to show that the framework achieves competitive metrics across all of the attacks (average true positive rate of 96.22% and an average true negative rate of 95.36%)

244. PoTrojan: powerful neural-level trojan designs in deep learning models

245. Hardware Trojan Attacks on Neural Networks

246. Spectral Signatures in Backdoor Attacks

     Summary

     • Identified a "spectral signatures" property of current backdoor attacks which allows the authors to use robust statistics to stop Trojan attacks
     • The "spectral signature" refers to a change in the covariance spectrum of learned feature representations that is left after a network is attacked. This can be detected by using singular value decomposition (SVD). SVD is used to identify which examples to remove from the training set. After these examples are removed the model is retrained on the cleaned dataset and is no longer Trojaned. The authors test this method on the CIFAR 10 image dataset.

247. Defending Neural Backdoors via Generative Distribution Modeling

248. Detecting Backdoor Attacks on Deep Neural Networks by Activation Clustering

     Summary

     • Proposes Activation Clustering approach to backdoor detection/ removal which analyzes the neural network activations for anomalies and works for both text and images
     • Activation Clustering uses dimensionality techniques (ICA, PCA) on the activations and then clusters them using k-means (k=2) along with a silhouette score metric to separate poisoned from clean clusters
     • Shows that Activation Clustering is successful on three different image/datasets (MNIST, LISA, Rotten Tomatoes) as well as in settings where multiple Trojans are inserted and classes are multi-modal

249. Model-Reuse Attacks on Deep Learning Systems

250. How To Backdoor Federated Learning

251. Trojaning Attack on Neural Networks

252. Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networks

     Summary

     • Proposes neural network poisoning attack that uses "clean labels" which do not require the adversary to mislabel training inputs
     • The paper also presents a optimization based method for generating their poisoning attacks and provides a watermarking strategy for end-to-end attacks that improves the poisoning reliability
     • Authors demonstrate their method by using generated poisoned frog images from the CIFAR dataset to manipulate different kinds of image classifiers

253. Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networks

     Summary

     • Investigate two potential detection methods for backdoor attacks (Fine-tuning and pruning). They find both are insufficient on their own and thus propose a combined detection method which they call "Fine-Pruning"
     • Authors go on to show that on three backdoor techniques "Fine-Pruning" is able to eliminate or reduce Trojans on datasets in the traffic sign, speech, and face recognition domains

254. Technical Report: When Does Machine Learning FAIL? Generalized Transferability for Evasion and Poisoning Attacks

255. Backdoor Embedding in Convolutional Neural Network Models via Invisible Perturbation

256. Hu-Fu: Hardware and Software Collaborative Attack Framework against Neural Networks

257. Attack Strength vs. Detectability Dilemma in Adversarial Machine Learning

258. Data Poisoning Attacks in Contextual Bandits

259. BEBP: An Poisoning Method Against Machine Learning Based IDSs

260. Generative Poisoning Attack Method Against Neural Networks

261. BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain

     Summary

     • Introduce Trojan Attacks— a type of attack where an adversary can create a maliciously trained network (a backdoored neural network, or a BadNet) that has state-of-the-art performance on the user’s training and validation samples, but behaves badly on specific attacker-chosen inputs
     • Demonstrate backdoors in a more realistic scenario by creating a U.S. street sign classifier that identifies stop signs as speed limits when a special sticker is added to the stop sign

262. Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization

263. Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning

264. Neural Trojans

265. Towards Poisoning of Deep Learning Algorithms with Back-gradient Optimization

266. Certified defenses for data poisoning attacks

267. Data Poisoning Attacks on Factorization-Based Collaborative Filtering

268. Data poisoning attacks against autoregressive models

269. Using machine teaching to identify optimal training-set attacks on machine learners

270. Poisoning Attacks against Support Vector Machines

271. Backdoor Attacks against Learning Systems

272. Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning

273. Antidote: Understanding and defending against poisoning of anomaly detectors