GARD cuts CLIP visual backdoor success below 1%
The post-hoc defense covers six visual backdoor attacks on two CLIP backbones and needs no extra inference compute.
Stories per day · 30 days
2 RSS
Relatedbackdoor2contrastive-learning2multimodal2ai-security1biomedclip1breast-imaging1celeb-df1clap1computer-vision1deepfake1differential-privacy1dinov21domain-generalization1embeddings1faceforensics1federated-learning1image-editing1imagenet1
17 stories, newest first · one card per story
Stories tagged #clip
rPPG deepfake detectors track high-frequency artifacts, not pulse
Chrominance traces hit AUC 0.998 on talking-face forgeries, and Celeb-DF splits leak identities into the test set.
SIFR fuses CLIP and DINOv2 spectrally for domain-general segmentation
The method filters domain-sensitive responses at fusion and stabilizes multi-scale decoder features before dense prediction.
PREEN predicts editor presets for exemplar color transfer
The model regresses 29 smartphone-editor sliders, runs at sub-second on-device latency, and leads PST50 fidelity metrics.
Federated breast-imaging models may be confounded by acquisition year
Normal-versus-sick ROC-AUC was 0.940-0.990, but year alone scored 0.974 on the same partition.
Simplex embeddings cut the multimodal modality gap by 97–99%
Softmax adapters and Total Variation similarity match or beat unconstrained retrieval on five benchmarks.
Clean Data Can Still Carry Backdoors: Support-Persistent Backdoors for Model Reuse
DAR plants support-persistent backdoors that stay above 96 percent success after clean-data distillation.
Chaos in the Text: Revealing the Modality Preference in Mixed-Modality Retrievers
Mixed-modality retrievers favor text over images; Trident reduces that bias.
Mu-DisCoCat: A Variational Pipeline for Compositional Generalization on Quantum Processors
Mu-DisCoCat uses variational quantum circuits to improve compositional generalization on noisy near-term quantum hardware.
Rethinking Long-Video Efficiency: A Joint Allocation Perspective on Frames, Pixels, and Front-End Latency
LoHi pairs dense low-resolution video with sparse high-resolution frames, raising long-video accuracy and cutting decode latency.
Walking the Embedding Space: Datastore Extraction from Multimodal RAG
Researchers show a black-box attack can extract hundreds of private images from multimodal RAG datastores.
Region-Level Black-Box Defense Against Stealthy Embedding-Space Backdoors in CLIP
CLIPGuard localizes and inpaints CLIP embedding backdoors, cutting attack success to about 1% in black-box tests.
Can Pixels Alone Reveal Image Origin? Minimax Limits and Learnable Interfaces for Passive Provenance
Paper proves pixel-only image provenance has a minimax limit and shows public CLIP verifiers fail under edits.
Robot Visions: Breaking reCAPTCHA at Zero Cost and Zero Shot
Researchers defeat Google reCAPTCHA using free local models CLIP and OWLv2, achieving 92.6% per-session success at zero cost.
Anatomical Grounding and Leakage-Aware Multimodal Contrastive Learning for Alzheimer's Disease Classification from Structural MRI
Study of label leakage and anatomical grounding in multimodal MRI models for Alzheimer's staging shows cognitive-score fusion accuracy of 87.3% is leakage-driven.
GoDeep: Annotation-Free Open-Vocabulary 3D Scene Understanding via Language-Space Lifting
GoDeep achieves annotation-free open-vocabulary 3D segmentation by grounding structured image descriptions in language-only embeddings, outperforming CLIP-based lifting on out-of-vocabulary objects.
Harnessing CLIP and DINO: An Uncertainty-Aware Cascaded Fusion Network for Generalizable Deepfake Image Detection
UCF-Net fuses CLIP and DINO features with entropy-based uncertainty weighting to improve generalizable deepfake image detection across generators.