Class Prototype

Class prototypes, representative feature vectors for each class, are central to many machine learning approaches, particularly in scenarios with limited data or incremental learning. Current research focuses on developing robust and efficient methods for generating and utilizing these prototypes, including techniques that address privacy concerns in federated learning, optimize prototype placement for improved class separation, and mitigate catastrophic forgetting in incremental learning settings. These advancements are crucial for improving the performance and adaptability of machine learning models in various applications, such as image classification, semantic segmentation, and domain adaptation, especially when dealing with data scarcity or evolving class distributions.

Papers

October 14, 2024

Improve Meta-learning for Few-Shot Text Classification with All You Can Acquire from the Tasks
Xinyue Liu, Yunlong Gao, Linlin Zong, Bo Xu
Pre Trained Language Model Text Classification Meta Learning Metric Learning New Task Short Text Class Prototype

September 12, 2024

FedHide: Federated Learning by Hiding in the Neighbors
Hyunsin Park, Sungrack Yun
Thy Neighbor Gradient Inversion Attack Leakage Attack Class Prototype Weapon of Mass Destruction Proxy

July 10, 2024

A Coding-Theoretic Analysis of Hyperspherical Prototypical Learning Geometry
Martin Lindström, Borja Rodríguez-Gálvez, Ragnar Thobaben, Mikael Skoglund
Representation Learning Information Theoretic Latent Prototype Hyperspherical Learning Class Prototype Multi Prototype

May 29, 2024

Organizing Background to Explore Latent Classes for Incremental Few-shot Semantic Segmentation
Lianlei Shan, Wenzhang Zhou, Wei Li, Xingyu Ding
Background Knowledge Pre Trained Segmentation Model Latent Class Class Prototype Novel Class Prototype Network

July 20, 2023

Feed-Forward Source-Free Domain Adaptation via Class Prototypes
Ondrej Bohdal, Da Li, Timothy Hospedales
Domain Adaptation Pre Trained Model Domain Shift Back Propagation Source Free Domain Adaptation Class Prototype

July 5, 2023

S3C: Self-Supervised Stochastic Classifiers for Few-Shot Class-Incremental Learning
Jayateja Kalla, Soma Biswas
Class Incremental Learning Unseen Class Class Prototype

July 4, 2023

ProtoDiffusion: Classifier-Free Diffusion Guidance with Prototype Learning
Gulcin Baykal, Halil Faruk Karagoz, Taha Binhuraib, Gozde Unal
Diffusion Model Generative Model Diffusion Guidance Class Prototype

March 20, 2023

Non-Exemplar Online Class-incremental Continual Learning via Dual-prototype Self-augment and Refinement
Fushuo Huo, Wenchao Xu, Jingcai Guo, Haozhao Wang, Yunfeng Fan, Song Guo
Class Incremental Learning Self Augmentation Class Incremental Continual Learning Class Prototype Prototype Highlighting Current Challenge

December 6, 2022

Tackling Data Heterogeneity in Federated Learning with Class Prototypes
Yutong Dai, Zeyuan Chen, Junnan Li, Shelby Heinecke, Lichao Sun, Ran Xu
Strong Generalization Class Imbalance Data Heterogeneity Global Model Local Model Class Prototype

October 4, 2022

COARSE3D: Class-Prototypes for Contrastive Learning in Weakly-Supervised 3D Point Cloud Segmentation
Rong Li, Anh-Quan Cao, Raoul de Charette
Contrastive Learning 3D Segmentation Large Scale 3D Weakly Supervised Point Cloud Class Prototype Model Agnostic Contrastive

July 18, 2022

Class-incremental Novel Class Discovery
Subhankar Roy, Mingxuan Liu, Zhun Zhong, Nicu Sebe, Elisa Ricci
Knowledge Distillation Novel Class Discovery Class Prototype Rehearsal Based Continual Learning

March 31, 2022

Few-Shot Class-Incremental Learning by Sampling Multi-Phase Tasks
Da-Wei Zhou, Han-Jia Ye, Liang Ma, Di Xie, Shiliang Pu, De-Chuan Zhan
Supervised Learning Class Incremental Learning Generalizable Feature Shot Class Class Prototype Multi Stage Task Learning Multi Phase Incremental Task