out-of-distribution samples
Out-of-distribution samples are data points that originate from a different distribution than the one on which the model was trained. Identifying and handling these samples is crucial for maintaining model performance in real-world applications.
- Bayesian Concept Bottleneck Models with LLM Priors
- CLIPTTA: Robust Contrastive Vision-Language Test-Time Adaptation
- Compositional Discrete Latent Code for High Fidelity, Productive Diffusion Models
- GOOD: Training-Free Guided Diffusion Sampling for Out-of-Distribution Detection
- OOD-Barrier: Build a Middle-Barrier for Open-Set Single-Image Test Time Adaptation via Vision Language Models
- Versatile Transferable Unlearnable Example Generator