no code implementations • 28 Feb 2023 • Chong Fu, Xuhong Zhang, Shouling Ji, Ting Wang, Peng Lin, Yanghe Feng, Jianwei Yin
Thus, in this paper, we propose FreeEagle, the first data-free backdoor detection method that can effectively detect complex backdoor attacks on deep neural networks, without relying on the access to any clean samples or samples with the trigger.
no code implementations • 12 Feb 2023 • Lujia Shen, Xuhong Zhang, Shouling Ji, Yuwen Pu, Chunpeng Ge, Xing Yang, Yanghe Feng
TextDefense differs from previous approaches, where it utilizes the target model for detection and thus is attack type agnostic.
no code implementations • 1 Dec 2022 • Pengyu Qiu, Xuhong Zhang, Shouling Ji, Yuwen Pu, Ting Wang
Vertical federated learning is a trending solution for multi-party collaboration in training machine learning models.
no code implementations • 1 Dec 2022 • Pengyu Qiu, Xuhong Zhang, Shouling Ji, Changjiang Li, Yuwen Pu, Xing Yang, Ting Wang
Vertical federated learning (VFL) is an emerging paradigm that enables collaborators to build machine learning models together in a distributed fashion.
no code implementations • 21 Oct 2022 • Ren Pang, Changjiang Li, Zhaohan Xi, Shouling Ji, Ting Wang
This paper asks the intriguing question: is it possible to exploit neural architecture search (NAS) as a new attack vector to launch previously improbable attacks?
no code implementations • 13 Oct 2022 • Changjiang Li, Ren Pang, Zhaohan Xi, Tianyu Du, Shouling Ji, Yuan YAO, Ting Wang
We explore this question in the context of trojan attacks by showing that SSL is comparably vulnerable as supervised learning to trojan attacks.
2 code implementations • USENIX Security 22 2022 • Chong Fu, Xuhong Zhang, Shouling Ji, Jinyin Chen, Jingzheng Wu, Shanqing Guo, Jun Zhou, Alex X. Liu, Ting Wang
However, we discover that the bottom model structure and the gradient update mechanism of VFL can be exploited by a malicious participant to gain the power to infer the privately owned labels.
no code implementations • 27 Sep 2022 • Zhaohan Xi, Ren Pang, Changjiang Li, Tianyu Du, Shouling Ji, Fenglong Ma, Ting Wang
(ii) It supports complex logical queries with varying relation and view constraints (e. g., with complex topology and/or from multiple views); (iii) It scales up to KGs of large sizes (e. g., millions of facts) and fine-granular views (e. g., dozens of views); (iv) It generalizes to query structures and KG views that are unobserved during training.
no code implementations • 5 Sep 2022 • Yuyou Gan, Yuhao Mao, Xuhong Zhang, Shouling Ji, Yuwen Pu, Meng Han, Jianwei Yin, Ting Wang
Experiment results show that the MeTFA-smoothed explanation can significantly increase the robust faithfulness.
no code implementations • 25 May 2022 • Xiangshan Gao, Xingjun Ma, Jingyi Wang, Youcheng Sun, Bo Li, Shouling Ji, Peng Cheng, Jiming Chen
One desirable property for FL is the implementation of the right to be forgotten (RTBF), i. e., a leaving participant has the right to request to delete its private data from the global model.
no code implementations • 21 May 2022 • Yangkai Du, Tengfei Ma, Lingfei Wu, Yiming Wu, Xuhong Zhang, Bo Long, Shouling Ji
Towards real-world information extraction scenario, research of relation extraction is advancing to document-level relation extraction(DocRE).
Ranked #24 on Relation Extraction on DocRED
no code implementations • 7 Apr 2022 • Yuhao Mao, Chong Fu, Saizhuo Wang, Shouling Ji, Xuhong Zhang, Zhenguang Liu, Jun Zhou, Alex X. Liu, Raheem Beyah, Ting Wang
To bridge this critical gap, we conduct the first large-scale systematic empirical study of transfer attacks against major cloud-based MLaaS platforms, taking the components of a real transfer attack into account.
no code implementations • 13 Mar 2022 • Dayong Ye, Huiqiang Chen, Shuai Zhou, Tianqing Zhu, Wanlei Zhou, Shouling Ji
However, they may not mean that transfer learning models are impervious to model inversion attacks.
no code implementations • 22 Feb 2022 • Changjiang Li, Li Wang, Shouling Ji, Xuhong Zhang, Zhaohan Xi, Shanqing Guo, Ting Wang
Facial Liveness Verification (FLV) is widely used for identity authentication in many security-sensitive domains and offered as Platform-as-a-Service (PaaS) by leading cloud vendors.
1 code implementation • 21 Feb 2022 • Sihao Hu, Yi Cao, Yu Gong, Zhao Li, Yazheng Yang, Qingwen Liu, Shouling Ji
Specifically, we establish a heterogeneous graph that contains physical and semantic linkages to guide the feature transfer process from warmed-up video to cold-start videos.
1 code implementation • 30 Dec 2021 • Zhenguang Liu, Shuang Wu, Shuyuan Jin, Shouling Ji, Qi Liu, Shijian Lu, Li Cheng
One aspect that has been obviated so far, is the fact that how we represent the skeletal pose has a critical impact on the prediction results.
1 code implementation • 25 Dec 2021 • Haibin Zheng, Zhiqing Chen, Tianyu Du, Xuhong Zhang, Yao Cheng, Shouling Ji, Jingyi Wang, Yue Yu, Jinyin Chen
To overcome the challenges, we propose NeuronFair, a new DNN fairness testing framework that differs from previous work in several key aspects: (1) interpretable - it quantitatively interprets DNNs' fairness violations for the biased decision; (2) effective - it uses the interpretation results to guide the generation of more diverse instances in less time; (3) generic - it can handle both structured and unstructured data.
no code implementations • 24 Dec 2021 • Ruoxi Chen, Haibo Jin, Jinyin Chen, Haibin Zheng, Yue Yu, Shouling Ji
From the perspective of image feature space, some of them cannot reach satisfying results due to the shift of features.
1 code implementation • 23 Dec 2021 • Xiang Ling, Lingfei Wu, Jiangyu Zhang, Zhenqing Qu, Wei Deng, Xiang Chen, Yaguan Qian, Chunming Wu, Shouling Ji, Tianyue Luo, Jingzheng Wu, Yanjun Wu
Then, we conduct a comprehensive and systematic review to categorize the state-of-the-art adversarial attacks against PE malware detection, as well as corresponding defenses to increase the robustness of Windows PE malware detection.
no code implementations • 30 Oct 2021 • Lujia Shen, Shouling Ji, Xuhong Zhang, Jinfeng Li, Jing Chen, Jie Shi, Chengfang Fang, Jianwei Yin, Ting Wang
However, a pre-trained model with backdoor can be a severe threat to the applications.
1 code implementation • 12 Oct 2021 • Ren Pang, Zhaohan Xi, Shouling Ji, Xiapu Luo, Ting Wang
Neural Architecture Search (NAS) represents an emerging machine learning (ML) paradigm that automatically searches for models tailored to given tasks, which greatly simplifies the development of ML systems and propels the trend of ML democratization.
no code implementations • 9 Jul 2021 • Zuohui Chen, Renxuan Wang, Jingyang Xiang, Yue Yu, Xin Xia, Shouling Ji, Qi Xuan, Xiaoniu Yang
Deep Neural Networks (DNN) are known to be vulnerable to adversarial samples, the detection of which is crucial for the wide application of these DNN models.
1 code implementation • 17 Jun 2021 • Zhenguang Liu, Peng Qian, Xiang Wang, Lei Zhu, Qinming He, Shouling Ji
In this paper, we explore combining deep learning with expert patterns in an explainable fashion.
1 code implementation • Findings (EMNLP) 2021 • Yangkai Du, Tengfei Ma, Lingfei Wu, Fangli Xu, Xuhong Zhang, Bo Long, Shouling Ji
Unlike vision tasks, the data augmentation method for contrastive learning has not been investigated sufficiently in language tasks.
1 code implementation • 6 Apr 2021 • Jianfeng Dong, Zhe Ma, Xiaofeng Mao, Xun Yang, Yuan He, Richang Hong, Shouling Ji
In this similarity paradigm, one should pay more attention to the similarity in terms of a specific design/attribute between fashion items.
1 code implementation • 25 Mar 2021 • Rui Wang, Zhihua Wei, Haoran Duan, Shouling Ji, Yang Long, Zhen Hong
Compared with hand-designed approaches, neural architecture search (NAS) appears as a practical technique in automating the manual architecture design process and has attracted increasing interest in spoken language processing tasks such as speaker recognition.
no code implementations • 17 Mar 2021 • Zhenguang Liu, Kedi Lyu, Shuang Wu, Haipeng Chen, Yanbin Hao, Shouling Ji
Our method is compelling in that it enables manipulable motion prediction across activity types and allows customization of the human movement in a variety of fine-grained ways.
1 code implementation • CVPR 2021 • Zhenguang Liu, Haoming Chen, Runyang Feng, Shuang Wu, Shouling Ji, Bailin Yang, Xun Wang
Multi-frame human pose estimation in complicated situations is challenging.
Ranked #1 on Multi-Person Pose Estimation on PoseTrack2017 (using extra training data)
no code implementations • 23 Feb 2021 • Jinfeng Li, Tianyu Du, Xiangyu Liu, Rong Zhang, Hui Xue, Shouling Ji
Extensive experiments on two real-world tasks show that AdvGraph exhibits better performance compared with previous work: (i) effective - it significantly strengthens the model robustness even under the adaptive attacks setting without negative impact on model performance over legitimate input; (ii) generic - its key component, i. e., the representation of connotative adversarial knowledge is task-agnostic, which can be reused in any Chinese-based NLP models without retraining; and (iii) efficient - it is a light-weight defense with sub-linear computational complexity, which can guarantee the efficiency required in practical scenarios.
1 code implementation • 18 Feb 2021 • Zhe Ma, Fenghao Liu, Jianfeng Dong, Xiaoye Qu, Yuan He, Shouling Ji
In this paper, we focus on the cross-modal similarity measurement, and propose a novel Hierarchical Similarity Learning (HSL) network.
no code implementations • 1 Feb 2021 • Yaguan Qian, Qiqi Shao, Tengteng Yao, Bin Wang, Shouling Ji, Shaoning Zeng, Zhaoquan Gu, Wassim Swaileh
Adversarial training is wildly considered as one of the most effective way to defend against adversarial examples.
no code implementations • 22 Jan 2021 • Xinyang Zhang, Ren Pang, Shouling Ji, Fenglong Ma, Ting Wang
Providing explanations for deep neural networks (DNNs) is essential for their use in domains wherein the interpretability of decisions is a critical prerequisite.
no code implementations • 1 Jan 2021 • Xiang Ling, Lingfei Wu, Saizhuo Wang, Tengfei Ma, Fangli Xu, Alex X. Liu, Chunming Wu, Shouling Ji
The proposed MGMN model consists of a node-graph matching network for effectively learning cross-level interactions between nodes of a graph and the other whole graph, and a siamese graph neural network to learn global-level interactions between two graphs.
1 code implementation • 16 Dec 2020 • Ren Pang, Zheng Zhang, Xiangshan Gao, Zhaohan Xi, Shouling Ji, Peng Cheng, Xiapu Luo, Ting Wang
To bridge this gap, we design and implement TROJANZOO, the first open-source platform for evaluating neural backdoor attacks/defenses in a unified, holistic, and practical manner.
no code implementations • 2 Dec 2020 • Yaguan Qian, Jiamin Wang, Bin Wang, Shaoning Zeng, Zhaoquan Gu, Shouling Ji, Wassim Swaileh
With this soft mask, we develop a new loss function with inverse temperature to search for optimal perturbations in CFR.
no code implementations • 25 Oct 2020 • Hanlu Wu, Tengfei Ma, Lingfei Wu, Shouling Ji
Besides, we exploit the unknown latent interaction between the same type of nodes (workers or tasks) by adding a homogeneous attention layer in the graph neural networks.
no code implementations • 24 Oct 2020 • Xiang Ling, Lingfei Wu, Saizhuo Wang, Gaoning Pan, Tengfei Ma, Fangli Xu, Alex X. Liu, Chunming Wu, Shouling Ji
To this end, we first represent both natural language query texts and programming language code snippets with the unified graph-structured data, and then use the proposed graph matching and searching model to retrieve the best matching code snippet.
1 code implementation • 5 Oct 2020 • Yuwei Li, Shouling Ji, Yuan Chen, Sizhuang Liang, Wei-Han Lee, Yueyao Chen, Chenyang Lyu, Chunming Wu, Raheem Beyah, Peng Cheng, Kangjie Lu, Ting Wang
We hope that our findings can shed light on reliable fuzzing evaluation, so that we can discover promising fuzzing primitives to effectively facilitate fuzzer designs in the future.
Cryptography and Security
1 code implementation • EMNLP 2020 • Hanlu Wu, Tengfei Ma, Lingfei Wu, Tariro Manyumwa, Shouling Ji
Experiments on Newsroom and CNN/Daily Mail demonstrate that our new evaluation method outperforms other metrics even without reference summaries.
1 code implementation • 1 Aug 2020 • Xinyang Zhang, Zheng Zhang, Shouling Ji, Ting Wang
Recent years have witnessed the emergence of a new paradigm of building natural language processing (NLP) systems: general-purpose, pre-trained language models (LMs) are composed with simple downstream models and fine-tuned for a variety of NLP tasks.
1 code implementation • 8 Jul 2020 • Xiang Ling, Lingfei Wu, Saizhuo Wang, Tengfei Ma, Fangli Xu, Alex X. Liu, Chunming Wu, Shouling Ji
In particular, the proposed MGMN consists of a node-graph matching network for effectively learning cross-level interactions between each node of one graph and the other whole graph, and a siamese graph neural network to learn global-level interactions between two input graphs.
1 code implementation • 21 Jun 2020 • Zhaohan Xi, Ren Pang, Shouling Ji, Ting Wang
One intriguing property of deep neural networks (DNNs) is their inherent vulnerability to backdoor attacks -- a trojan model responds to trigger-embedded inputs in a highly predictable manner while functioning normally otherwise.
1 code implementation • 16 Jun 2020 • Ren Pang, Xinyang Zhang, Shouling Ji, Xiapu Luo, Ting Wang
Deep neural networks (DNNs) are inherently susceptible to adversarial attacks even under black-box settings, in which the adversary only has query access to the target models.
1 code implementation • 7 Feb 2020 • Zhe Ma, Jianfeng Dong, Yao Zhang, Zhongzi Long, Yuan He, Hui Xue, Shouling Ji
This paper strives to learn fine-grained fashion similarity.
no code implementations • 25 Nov 2019 • Lingfei Wu, Ian En-Hsu Yen, Siyu Huo, Liang Zhao, Kun Xu, Liang Ma, Shouling Ji, Charu Aggarwal
In this paper, we present a new class of global string kernels that aims to (i) discover global properties hidden in the strings through global alignments, (ii) maintain positive-definiteness of the kernel, without introducing a diagonal dominant kernel matrix, and (iii) have a training cost linear with respect to not only the length of the string but also the number of training string samples.
1 code implementation • 5 Nov 2019 • Ren Pang, Hua Shen, Xinyang Zhang, Shouling Ji, Yevgeniy Vorobeychik, Xiapu Luo, Alex Liu, Ting Wang
Specifically, (i) we develop a new attack model that jointly optimizes adversarial inputs and poisoned models; (ii) with both analytical and empirical evidence, we reveal that there exist intriguing "mutual reinforcement" effects between the two attack vectors -- leveraging one vector significantly amplifies the effectiveness of the other; (iii) we demonstrate that such effects enable a large design spectrum for the adversary to enhance the existing attacks that exploit both vectors (e. g., backdoor attacks), such as maximizing the attack evasiveness with respect to various detection methods; (iv) finally, we discuss potential countermeasures against such optimized attacks and their technical challenges, pointing to several promising research directions.
no code implementations • 25 Sep 2019 • Xiang Ling, Lingfei Wu, Saizhuo Wang, Tengfei Ma, Fangli Xu, Chunming Wu, Shouling Ji
The proposed HGMN model consists of a multi-perspective node-graph matching network for effectively learning cross-level interactions between parts of a graph and a whole graph, and a siamese graph neural network for learning global-level interactions between two graphs.
no code implementations • 7 Aug 2019 • Zhikang Zou, Yu Cheng, Xiaoye Qu, Shouling Ji, Xiaoxiao Guo, Pan Zhou
ACM-CNN consists of three types of modules: a coarse network, a fine network, and a smooth network.
no code implementations • ICLR 2019 • Xinyang Zhang, Yifan Huang, Chanh Nguyen, Shouling Ji, Ting Wang
On the possibility side, we show that it is still feasible to construct adversarial training methods to significantly improve the resilience of networks against adversarial inputs over empirical datasets.
no code implementations • 23 Jan 2019 • Tianyu Du, Shouling Ji, Jinfeng Li, Qinchen Gu, Ting Wang, Raheem Beyah
Despite their immense popularity, deep learning-based acoustic systems are inherently vulnerable to adversarial attacks, wherein maliciously crafted audios trigger target systems to misbehave.
Cryptography and Security
1 code implementation • 10 Jan 2019 • Jian-hai Chen, Deshi Ye, Shouling Ji, Qinming He, Yang Xiang, Zhenguang Liu
Next, we prove that our mechanism is an FPTAS, i. e., it can be approximated within $1 + \epsilon$ for any given $\epsilon > 0$, while the running time of our mechanism is polynomial in $n$ and $1/\epsilon$, where $n$ is the number of tenants in the datacenter.
Computer Science and Game Theory
no code implementations • 4 Jan 2019 • Yuwei Li, Shouling Ji, Chenyang Lv, Yu-An Chen, Jian-hai Chen, Qinchen Gu, Chunming Wu
Given a binary program to V-Fuzz, the vulnerability prediction model will give a prior estimation on which parts of the software are more likely to be vulnerable.
Cryptography and Security
no code implementations • 4 Jan 2019 • Xurong Li, Shouling Ji, Meng Han, Juntao Ji, Zhenyu Ren, Yushan Liu, Chunming Wu
Through the comprehensive evaluations on five major cloud platforms: AWS, Azure, Google Cloud, Baidu Cloud, and Alibaba Cloud, we demonstrate that our image processing based attacks can reach a success rate of approximately 100%, and the semantic segmentation based attacks have a success rate over 90% among different detection services, such as violence, politician, and pornography detection.
1 code implementation • 13 Dec 2018 • Jinfeng Li, Shouling Ji, Tianyu Du, Bo Li, Ting Wang
Deep Learning-based Text Understanding (DLTU) is the backbone technique behind various applications, including question answering, machine translation, and text classification.
no code implementations • 3 Dec 2018 • Xinyang Zhang, Ningfei Wang, Hua Shen, Shouling Ji, Xiapu Luo, Ting Wang
The improved interpretability is believed to offer a sense of security by involving human in the decision-making process.
no code implementations • 2 Dec 2018 • Yujie Ji, Xinyang Zhang, Shouling Ji, Xiapu Luo, Ting Wang
By empirically studying four deep learning systems (including both individual and ensemble systems) used in skin cancer screening, speech recognition, face verification, and autonomous steering, we show that such attacks are (i) effective - the host systems misbehave on the targeted inputs as desired by the adversary with high probability, (ii) evasive - the malicious models function indistinguishably from their benign counterparts on non-targeted inputs, (iii) elastic - the malicious models remain effective regardless of various system design choices and tuning strategies, and (iv) easy - the adversary needs little prior knowledge about the data used for system tuning or inference.
Cryptography and Security
1 code implementation • CVPR 2019 • Jianfeng Dong, Xirong Li, Chaoxi Xu, Shouling Ji, Yuan He, Gang Yang, Xun Wang
This paper attacks the challenging problem of zero-example video retrieval.
no code implementations • 7 Jul 2018 • Chenyang Lyu, Shouling Ji, Yuwei Li, Junfeng Zhou, Jian-hai Chen, Jing Chen
In total, our system discovers more than twice unique crashes and 5, 040 extra unique paths than the existing best seed selection strategy for the evaluated 12 applications.
Cryptography and Security
2 code implementations • 5 Jan 2018 • Xinyang Zhang, Shouling Ji, Ting Wang
Privacy-preserving releasing of complex data (e. g., image, text, audio) represents a long-standing challenge for the data mining research community.