publications
2024
- Can We Trust the Unlabeled Target Data? Towards Backdoor Attack and Defense on Model AdaptationarXiv preprint arXiv:2401.06030, 2024
- ICMLRealistic Unsupervised CLIP Fine-tuning with Universal Entropy OptimizationIn Forty-first International Conference on Machine Learning , 2024
- IJCVLearning Spatiotemporal Inconsistency via Thumbnail Layout for Face Deepfake DetectionInternational Journal of Computer Vision, 2024
- ICLRA hard-to-beat baseline for training-free clip-based adaptationIn ICLR , 2024
2023
- ICCVAdaptguard: Defending against universal attacks for model adaptationIn Proceedings of the IEEE/CVF International Conference on Computer Vision , 2023
- NNProxymix: Proxy-based mixup training with label refinery for source-free domain adaptationNeural Networks, 2023
- ICCVWSelf-training solutions for the ICCV 2023 GeoNet ChallengearXiv preprint arXiv:2311.16843, 2023
- Unleashing the power of Neural Collapse for Transferability EstimationarXiv preprint arXiv:2310.05754, 2023
- Benchmarking test-time adaptation against distribution shifts in image classificationarXiv preprint arXiv:2307.03133, 2023