Generative prompt model for weakly supervised object localization

Y Zhao, Q Ye, W Wu, C Shen… - Proceedings of the IEEE …, 2023 - openaccess.thecvf.com
Weakly supervised object localization (WSOL) remains challenging when learning object
localization models from image category labels. Conventional methods that discriminatively …

Stealthy backdoor attack for code models

Z Yang, B Xu, JM Zhang, HJ Kang… - IEEE Transactions on …, 2024 - ieeexplore.ieee.org
Code models, such as CodeBERT and CodeT5, offer general-purpose representations of
code and play a vital role in supporting downstream automated software engineering tasks …

{PrivImage}: Differentially Private Synthetic Image Generation using Diffusion Models with {Semantic-Aware} Pretraining

K Li, C Gong, Z Li, Y Zhao, X Hou, T Wang - 33rd USENIX Security …, 2024 - usenix.org
Differential Privacy (DP) image data synthesis, which leverages the DP technique to
generate synthetic data to replace the sensitive data, allowing organizations to share and …

Continual learning for image segmentation with dynamic query

W Wu, Y Zhao, Z Li, L Shan, H Zhou… - IEEE Transactions on …, 2023 - ieeexplore.ieee.org
Image segmentation based on continual learning exhibits a critical drop of performance,
mainly due to catastrophic forgetting and background shift, as they are required to …

Towards fair machine learning software: Understanding and addressing model bias through counterfactual thinking

Z Wang, Y Zhou, M Qiu, I Haque, L Brown, Y He… - arXiv preprint arXiv …, 2023 - arxiv.org
The increasing use of Machine Learning (ML) software can lead to unfair and unethical
decisions, thus fairness bugs in software are becoming a growing concern. Addressing …

Flowtext: Synthesizing realistic scene text video with optical flow estimation

Y Zhao, W Wu, Z Li, J Li, W Wang - 2023 IEEE International …, 2023 - ieeexplore.ieee.org
Current video text spotting methods can achieve preferable performance, powered with
sufficient labeled training data. However, labeling data manually is time-consuming and …

What do users ask in open-source AI repositories? An empirical study of GitHub issues

Z Yang, C Wang, J Shi, T Hoang… - 2023 IEEE/ACM 20th …, 2023 - ieeexplore.ieee.org
Artificial Intelligence (AI) systems, which benefit from the availability of large-scale datasets
and increasing computational power, have become effective solutions to various critical …

MORTAR: A Model-based Runtime Action Repair Framework for AI-enabled Cyber-Physical Systems

R Wang, Z Zhou, J Song, X Xie, X Xie, L Ma - arXiv preprint arXiv …, 2024 - arxiv.org
Cyber-Physical Systems (CPSs) are increasingly prevalent across various industrial and
daily-life domains, with applications ranging from robotic operations to autonomous driving …

Prioritizing speech test cases

Z Yang, J Shi, MH Asyrofi, B Xu, X Zhou… - ACM Transactions on …, 2023 - dl.acm.org
As automated speech recognition (ASR) systems gain widespread acceptance, there is a
pressing need to rigorously test and enhance their performance. Nonetheless, the process …

Meticulously selecting 1% of the dataset for pre-training! generating differentially private images data with semantics query

K Li, C Gong, Z Li, Y Zhao, X Hou, T Wang - arXiv preprint arXiv …, 2023 - arxiv.org
Differential Privacy (DP) image data synthesis, which leverages the DP technique to
generate synthetic data to replace the sensitive data, allowing organizations to share and …