🏢 School of Computer Science and Engineering, Southeast University

Vision-language models effectively detect noisy labels, improving image classification accuracy with DEFT.

SLAN: A novel approach for multi-label open-set recognition, enriching sub-labeling info using structural data to identify unknown labels.

MIPLMA, a novel algorithm, enhances multi-instance partial-label learning by dynamically adjusting margins for attention scores and predicted probabilities, leading to superior performance.

Linearly decompose & recompose Vision Transformers to create diverse-scale models efficiently, reducing computational costs & improving flexibility for various applications.

LeTs: Learnable Transformation efficiently initializes variable-sized Vision Transformers by learning adaptable transformations from a compact learngene module, outperforming from-scratch training.

CCL, a novel probabilistic framework, uses continuous contrastive learning to excel in long-tailed semi-supervised recognition, surpassing prior state-of-the-art methods by over 4%.

Cluster-Learngene efficiently initializes elastic-scale Vision Transformers by adaptively clustering and inheriting key modules from a large ancestry model, saving resources and boosting downstream ta…

A novel motion-aware spatio-temporal graph model surpasses existing methods in video salient object ranking by jointly optimizing multi-scale spatial and temporal features, thus accurately prioritizin…