34 citations · 130 across the 35 of their papers we have counts for
37 papers
Unifying Perplexing Behaviors in Modified BP Attributions through Alignment Perspective
Guanhua Zheng, Jitao Sang, Changsheng Xu
Attributions aim to identify input pixels that are relevant to the decision-making process. A popular approach involves using modified backpropagation (BP) rules to reverse decisio…
Benign Shortcut for Debiasing: Fair Visual Recognition via Intervention with Shortcut Features
Yi Zhang, Jitao Sang, Junyang Wang +2
Machine learning models often learn to make predictions that rely on sensitive social attributes like gender and race, which poses significant fairness risks, especially in societa…
You talk what you read: Understanding News Comment Behavior by Dispositional and Situational Attribution
Yuhang Wang, Yuxiang Zhang, Dongyuan Lu +1
Many news comment mining studies are based on the assumption that comment is explicitly linked to the corresponding news. In this paper, we observed that users' comments are also h…
Evaluation and Analysis of Hallucination in Large Vision-Language Models
Junyang Wang, Yiyang Zhou, Guohai Xu +9
Large Vision-Language Models (LVLMs) have recently achieved remarkable success. However, LVLMs are still plagued by the hallucination problem, which limits the practicality in many…
CValues: Measuring the Values of Chinese Large Language Models from Safety to Responsibility
Guohai Xu, Jiayi Liu, Ming Yan +11
With the rapid evolution of large language models (LLMs), there is a growing concern that they may pose risks or have negative social impacts. Therefore, evaluation of human values…
MF-CLIP: Leveraging CLIP as Surrogate Models for No-box Adversarial Attacks
Jiaming Zhang, Lingyu Qiu, Qi Yi +4
The vulnerability of Deep Neural Networks (DNNs) to adversarial attacks poses a significant challenge to their deployment in safety-critical applications. While extensive research…
Towards Alleviating the Object Bias in Prompt Tuning-based Factual Knowledge Extraction
Yuhang Wang, Dongyuan Lu, Chao Kong +1
Many works employed prompt tuning methods to automatically optimize prompt queries and extract the factual knowledge stored in Pretrained Language Models. In this paper, we observe…
Towards Black-box Adversarial Example Detection: A Data Reconstruction-based Method
Yifei Gao, Zhiyu Lin, Yunfan Yang +1
Adversarial example detection is known to be an effective adversarial defense method. Black-box attack, which is a more realistic threat and has led to various black-box adversaria…
Echoes: Unsupervised Debiasing via Pseudo-bias Labeling in an Echo Chamber
Rui Hu, Yahan Tu, Jitao Sang
Neural networks often learn spurious correlations when exposed to biased training data, leading to poor performance on out-of-distribution data. A biased dataset can be divided, ac…
Improved Visual Fine-tuning with Natural Language Supervision
Junyang Wang, Yuanhong Xu, Juhua Hu +3
Fine-tuning a visual pre-trained model can leverage the semantic information from large-scale pre-training data and mitigate the over-fitting problem on downstream vision tasks wit…
Backdoor for Debias: Mitigating Model Bias with Backdoor Attack-based Artificial Bias
Shangxi Wu, Qiuyang He, Jian Yu +1
With the swift advancement of deep learning, state-of-the-art algorithms have been utilized in various social situations. Nonetheless, some algorithms have been discovered to exhib…
Unlearnable Clusters: Towards Label-agnostic Unlearnable Examples
Jiaming Zhang, Xingjun Ma, Qi Yi +4
There is a growing interest in developing unlearnable examples (UEs) against visual privacy leaks on the Internet. UEs are training samples added with invisible but unlearnable noi…
Artificial Intelligence Security Competition (AISC)
Yinpeng Dong, Peng Chen, Senyou Deng +49
The security of artificial intelligence (AI) is an important research area towards safe, reliable, and trustworthy AI systems. To accelerate the research on AI security, the Artifi…
Zero-shot Image Captioning by Anchor-augmented Vision-Language Space Alignment
Junyang Wang, Yi Zhang, Ming Yan +2
CLIP (Contrastive Language-Image Pre-Training) has shown remarkable zero-shot transfer capabilities in cross-modal correlation tasks such as visual classification and image retriev…
Fair Visual Recognition via Intervention with Proxy Features
Yi Zhang, Jitao Sang, Junyang Wang
Deep learning models often learn to make predictions that rely on sensitive social attributes like gender and race, which poses significant fairness risks, especially in societal a…
FairCLIP: Social Bias Elimination based on Attribute Prototype Learning and Representation Neutralization
Junyang Wang, Yi Zhang, Jitao Sang
The Vision-Language Pre-training (VLP) models like CLIP have gained popularity in recent years. However, many works found that the social biases hidden in CLIP easily manifest in d…
Counterfactually Measuring and Eliminating Social Bias in Vision-Language Pre-training Models
Yi Zhang, Junyang Wang, Jitao Sang
Vision-Language Pre-training (VLP) models have achieved state-of-the-art performance in numerous cross-modal tasks. Since they are optimized to capture the statistical properties o…
Low-Mid Adversarial Perturbation against Unauthorized Face Recognition System
Jiaming Zhang, Qi Yi, Dongyuan Lu +1
In light of the growing concerns regarding the unauthorized use of facial recognition systems and its implications on individual privacy, the exploration of adversarial perturbatio…
Towards Adversarial Attack on Vision-Language Pre-training Models
Jiaming Zhang, Qi Yi, Jitao Sang
While vision-language pre-training model (VLP) has shown revolutionary improvements on various vision-language (V+L) tasks, the studies regarding its adversarial robustness remain…
Investigating and Explaining the Frequency Bias in Image Classification
Zhiyu Lin, Yifei Gao, Jitao Sang
CNNs exhibit many behaviors different from humans, one of which is the capability of employing high-frequency components. This paper discusses the frequency bias phenomenon in imag…
Non-generative Generalized Zero-shot Learning via Task-correlated Disentanglement and Controllable Samples Synthesis
Yaogong Feng, Xiaowen Huang, Pengbo Yang +2
Synthesizing pseudo samples is currently the most effective way to solve the Generalized Zero-Shot Learning (GZSL) problem. Most models achieve competitive performance but still su…
Debiasing Backdoor Attack: A Benign Application of Backdoor Attack in Eliminating Data Bias
Shangxi Wu, Qiuyang He, Yi Zhang +1
Backdoor attack is a new AI security risk that has emerged in recent years. Drawing on the previous research of adversarial attack, we argue that the backdoor attack has the potent…
Understanding and Testing Generalization of Deep Networks on Out-of-Distribution Data
Rui Hu, Jitao Sang, Jinqiang Wang +1
Deep network models perform excellently on In-Distribution (ID) data, but can significantly fail on Out-Of-Distribution (OOD) data. While developing methods focus on improving OOD…
Learning to Learn a Cold-start Sequential Recommender
Xiaowen Huang, Jitao Sang, Jian Yu +1
The cold-start recommendation is an urgent problem in contemporary online applications. It aims to provide users whose behaviors are literally sparse with as accurate recommendatio…