114 citations · 1k across the 73 of their papers we have counts for
15 papers · 1 filter
Robust Mixture-of-Expert Training for Convolutional Neural Networks
Yihua Zhang, Ruisi Cai, Tianlong Chen +6
Sparsely-gated Mixture of Expert (MoE), an emerging deep model architecture, has demonstrated a great promise to enable high-accuracy and ultra-efficient model inference. Despite t…
A Pilot Study of Query-Free Adversarial Attack against Stable Diffusion
Haomin Zhuang, Yihua Zhang, Sijia Liu
Despite the record-breaking performance in Text-to-Image (T2I) generation by Stable Diffusion, less research attention is paid to its adversarial robustness. In this work, we study…
Understanding and Improving Visual Prompting: A Label-Mapping Perspective
Aochuan Chen, Yuguang Yao, Pin-Yu Chen +2
We revisit and advance visual prompting (VP), an input prompting technique for vision tasks. VP can reprogram a fixed, pre-trained source model to accomplish downstream tasks in th…
Visual Prompting for Adversarial Robustness
Aochuan Chen, Peter Lorenz, Yuguang Yao +2
In this work, we leverage visual prompting (VP) to improve adversarial robustness of a fixed, pre-trained model at testing time. Compared to conventional adversarial defenses, VP a…
Reverse Engineering of Imperceptible Adversarial Image Perturbations
Yifan Gong, Yuguang Yao, Yize Li +4
It has been well recognized that neural network based image classifiers are easily fooled by images with tiny perturbations crafted by an adversary. There has been a vast volume of…
When Does Contrastive Learning Preserve Adversarial Robustness from Pretraining to Finetuning?
Lijie Fan, Sijia Liu, Pin-Yu Chen +2
Contrastive learning (CL) can learn generalizable feature representations and achieve the state-of-the-art performance of downstream tasks by finetuning a linear classifier on top…
Preserving Earlier Knowledge in Continual Learning with the Help of All Previous Feature Extractors
Zhuoyun Li, Changhong Zhong, Sijia Liu +2
Continual learning of new knowledge over time is one desirable capability for intelligent systems to recognize more and more classes of objects. Without or with very limited amount…
Achieving Real-Time LiDAR 3D Object Detection on a Mobile Device
Pu Zhao, Wei Niu, Geng Yuan +7
3D object detection is an important task, especially in the autonomous driving application domain. However, it is challenging to support the real-time performance with the limited…
Adversarial Robustness: From Self-Supervised Pre-Training to Fine-Tuning
Tianlong Chen, Sijia Liu, Shiyu Chang +3
Pretrained models from self-supervision are prevalently used in fine-tuning downstream tasks faster or for better accuracy. However, gaining robustness from pretraining is left une…
An Image Enhancing Pattern-based Sparsity for Real-time Inference on Mobile Devices
Xiaolong Ma, Wei Niu, Tianyun Zhang +8
Weight pruning has been widely acknowledged as a straightforward and effective method to eliminate redundancy in Deep Neural Networks (DNN), thereby achieving acceleration on vario…
Adversarial T-shirt! Evading Person Detectors in A Physical World
Kaidi Xu, Gaoyuan Zhang, Sijia Liu +6
It is known that deep neural networks (DNNs) are vulnerable to adversarial attacks. The so-called physical adversarial examples deceive DNN-based decisionmakers by attaching advers…
On the Design of Black-box Adversarial Examples by Leveraging Gradient-free Optimization and Operator Splitting Method
Pu Zhao, Sijia Liu, Pin-Yu Chen +4
Robust machine learning is currently one of the most prominent topics which could potentially help shaping a future of advanced AI platforms that not only perform well in average c…
Interpreting Adversarial Examples by Activation Promotion and Suppression
Kaidi Xu, Sijia Liu, Gaoyuan Zhang +5
It is widely known that convolutional neural networks (CNNs) are vulnerable to adversarial examples: images with imperceptible perturbations crafted to fool classifiers. However, i…
Adversarial Robustness vs Model Compression, or Both?
Shaokai Ye, Kaidi Xu, Sijia Liu +6
It is well known that deep neural networks (DNNs) are vulnerable to adversarial attacks, which are implemented by adding crafted perturbations onto benign examples. Min-max robust…
AutoZOOM: Autoencoder-based Zeroth Order Optimization Method for Attacking Black-box Neural Networks
Chun-Chen Tu, Paishun Ting, Pin-Yu Chen +5
Recent studies have shown that adversarial examples in state-of-the-art image classifiers trained by deep neural networks (DNN) can be easily generated when the target model is tra…