arxiv:2505.11049
Shengfang Zhai
zsf
AI & ML interests
Trustworthy AI, Generative Models, AI Privacy, Backdoor Attacks
Recent Activity
authored
a paper
5 days ago
NaviDet: Efficient Input-level Backdoor Detection on Text-to-Image
Synthesis via Neuron Activation Variation
authored
a paper
2 months ago
GuardReasoner-VL: Safeguarding VLMs via Reinforced Reasoning
published
a model
5 months ago
zsf/BadT2I_ObjBackdoor_motor2bike_u200b_4k_bsz64
Organizations
None yet
models
6
zsf/BadT2I_ObjBackdoor_dog2cat_u200b_8k_bsz16_augdata2k
Updated
•
7
zsf/BadT2I_ObjBackdoor_motor2bike_u200b_4k_bsz64
Updated
•
5
zsf/BadT2I_PixBackdoor_boya_u200b_2k_bsz16
Updated
•
6
zsf/BadT2I_StyBackdoor_blackandwhite_u200b_8k_bsz441
Updated
•
2
zsf/Unet_coco_50k141_ori_aug
Updated
•
2
zsf/Unet_coco_50k141_split1_aug
Updated