Accepted Papers

• Removing Backdoor Behaviors with Unlabeled Data (Oral)
  Lu Pang (Stony Brook University), Tao Sun (Stony Brook University), Haibin Ling (Stony Brook University), Chao Chen (Stony Brook University)

• Backdoor Attacks Against Transformers with Attention Enhancement (Oral)
  Weimin Lyu (Stony Brook University), Songzhu Zheng (Stony Brook University), Haibin Ling (Stony Brook University), Chao Chen (Stony Brook University)

• How to Backdoor Diffusion Models? (Oral)
  Sheng-Yen Chou (National Tsing Hua University & The Chinese University of Hong Kong), Pin-Yu Chen (IBM Research), Tsung-Yi Ho (The Chinese University of Hong Kong)

• BITE: Textual Backdoor Attacks with Iterative Trigger Injection (Spotlight Session #1, ID #1)
  Jun Yan (University of Southern California), Vansh Gupta (IIT Delhi), Xiang Ren (University of Southern California)

• Learning the Wrong Lessons: Inserting Trojans During Knowledge Distillation (Spotlight Session #1, ID #2)
  Leonard Tang (Harvard University), Tom Shlomi (Harvard University), Alexander Cai (Harvard University)

• Learning to Backdoor Federated Learning (Spotlight Session #1, ID #3)
  Henger Li (Tulane University), Chen Wu (Penn State University), Sencun Zhu (Penn State University), Zizhan Zheng (Tulane University)

• Secure Federated Learning against Model Poisoning Attacks via Client Filtering (Spotlight Session #1, ID #4)
  Duygu Nur Yaldiz (University of Southern California), Tuo Zhang (University of Southern California), Salman Avestimehr (University of Southern California)

• Unlearning Backdoor Attacks in Federated Learning (Spotlight Session #1, ID #5)
  Chen Wu (The Pennsylvania State University), Sencun Zhu (The Pennsylvania State University), Prasenjit Mitra (The Pennsylvania State University)

• Rethinking the Necessity of Labels in Backdoor Removal (Spotlight Session #1, ID #6)
  Zidi Xiong (University of Illinois Urbana-Champaign), Dongxian Wu (The University of Tokyo), Yifei Wang (Peking University), Yisen Wang (Peking University)

• BackdoorBox: A Python Toolbox for Backdoor Learning (Spotlight Session #2, ID #1)
  Yiming Li (Tsinghua University), Mengxi Ya (Tsinghua University), Yang Bai (Tsinghua University), Yong Jiang (Tsinghua University), Shu-Tao Xia (Tsinghua University)

• On the Existence of a Trojaned Twin Model (Spotlight Session #2, ID #2)
  Songzhu Zheng (Morgan Stanley), Yikai Zhang (Morgan Stanley), Lu Pang (Stony Brook University), Weimin Lyu (Stony Brook University), Mayank Goswami (City University of New York), Anderson Schneider (Morgan Stanley), Yuriy Nevmyvaka (Morgan Stanley), Haibin Ling (Stony Brook University), Chao Chen (Stony Brook University)

• DABS: Data-Agnostic Backdoor attack at the Server in Federated Learning (Spotlight Session #2, ID #3)
  Wenqiang Sun (The Hong Kong University of Science and Technology), Sen Li (The Hong Kong University of Science and Technology), Yuchang Sun (The Hong Kong University of Science and Technology), Jun Zhang (The Hong Kong University of Science and Technology)

• Exploring Vulnerabilities of Semi-Supervised Learning to Simple Backdoor Attacks (Spotlight Session #2, ID #4)
  Marissa Catherine Connor (Embedded Intelligence), Vincent Emanuele (Embedded Intelligence)

• Augmentation Backdoors (Spotlight Session #2, ID #5)
  Joseph Rance (University of Cambridge), Yiren Zhao (University of Cambridge & Imperial College London), Ilia Shumailov (University of Oxford), Robert D. Mullins (University of Cambridge)

• Salient Conditional Diffusion for Backdoors (Spotlight Session #2, ID #6)
  Brandon B May (STR), Norman Joseph Tatro (STR), Piyush Kumar (STR), Nathan Shnidman (STR)