5 papers
PANICL: Mitigating Over-Reliance on Single Prompt in Visual In-Context Learning
Jiahao Zhang, Bowen Wang, Hong Liu +2
Visual In-Context Learning (VICL) uses input-output image pairs, referred to as in-context pairs (or examples), as prompts alongside query images to guide models in performing dive…
QuMAB: Query-based Multi-Annotator Behavior Modeling with Reliability under Sparse Labels
Liyun Zhang, Zheng Lian, Hong Liu +2
Multi-annotator learning traditionally aggregates diverse annotations to approximate a single ground truth, treating disagreements as noise. However, this paradigm faces fundamenta…
SimLabel: Similarity-Weighted Iterative Framework for Multi-annotator Learning with Missing Annotations
Liyun Zhang, Zheng Lian, Hong Liu +2
Multi-annotator learning (MAL) aims to model annotator-specific labeling patterns. However, existing methods face a critical challenge: they simply skip updating annotator-specific…
QuMATL: Query-based Multi-annotator Tendency Learning
Liyun Zhang, Zheng Lian, Hong Liu +2
Different annotators often assign different labels to the same sample due to backgrounds or preferences, and such labeling patterns are referred to as tendency. In multi-annotator…
E-InMeMo: Enhanced Prompting for Visual In-Context Learning
Jiahao Zhang, Bowen Wang, Hong Liu +3
Large-scale models trained on extensive datasets have become the standard due to their strong generalizability across diverse tasks. In-context learning (ICL), widely used in natur…