Class Data

Class data research focuses on improving machine learning models' ability to handle situations where training data is incomplete or imbalanced, particularly concerning unseen classes. Current research emphasizes developing novel architectures, such as dual teacher-student models and gating networks, and algorithms like knowledge distillation and online placebo selection, to address challenges like catastrophic forgetting and class mismatch in semi-supervised, incremental, and zero-shot learning scenarios. These advancements aim to enhance model robustness and generalization, impacting various applications including medical image analysis and object recognition where data scarcity or continuous class evolution is common. The ultimate goal is to create more reliable and adaptable machine learning systems.

Papers

May 25, 2024

Diverse Teacher-Students for Deep Safe Semi-Supervised Learning under Class Mismatch
Qikai Wang, Rundong He, Yongshun Gong, Chunxiao Ren, Haoliang Sun, Xiaoshui Huang, Yilong Yin
Semi Supervised Learning Semi Supervised Unseen Class Unlabeled Dataset Class Sample Class Data

October 24, 2023

Wakening Past Concepts without Past Data: Class-Incremental Learning from Online Placebos
Yaoyao Liu, Yingying Li, Bernt Schiele, Qianru Sun
Knowledge Distillation Class Incremental Learning Unseen Entity Historical Data Class Data Placebo Effect

July 18, 2023

Class-relation Knowledge Distillation for Novel Class Discovery
Peiyan Gu, Chuyu Zhang, Ruijie Xu, Xuming He
Novel Class Discovery Relation Modeling Class Data

June 26, 2022

Class Impression for Data-free Incremental Learning
Sana Ayromlou, Purang Abolmaesumi, Teresa Tsang, Xiaoxiao Li
Training Data Incremental Learning Incremental Learning Framework Class Conditional Impression Reappearing Class Data

March 17, 2022

Semantic-diversity transfer network for generalized zero-shot learning via inner disagreement based OOD detector
Bo Liu, Qiulei Dong, Zhanyi Hu
Zero Shot Learning Visual Semantic Generalized Zero Shot Learning Diversity Regularizer Human Disagreement Class Data OOD Detector

March 8, 2022

A Gating Model for Bias Calibration in Generalized Zero-shot Learning
Gukyeong Kwon, Ghassan AlRegib
Supervised Autoencoder Generalized Zero Shot Learning Gating Mechanism Bias Calibration Class Data