Understanding imbalanced semantic segmentation through neural collapse

Z Zhong, J Cui, Y Yang, X Wu, X Qi… - Proceedings of the …, 2023 - openaccess.thecvf.com
A recent study has shown a phenomenon called neural collapse in that the within-class
means of features and the classifier weight vectors converge to the vertices of a simplex …

Imbalance trouble: Revisiting neural-collapse geometry

C Thrampoulidis, GR Kini… - Advances in Neural …, 2022 - proceedings.neurips.cc
Neural Collapse refers to the remarkable structural properties characterizing the geometry of
class embeddings and classifier weights, found by deep nets when trained beyond zero …

Neural collapse with normalized features: A geometric analysis over the riemannian manifold

C Yaras, P Wang, Z Zhu… - Advances in neural …, 2022 - proceedings.neurips.cc
When training overparameterized deep networks for classification tasks, it has been widely
observed that the learned features exhibit a so-called" neural collapse'" phenomenon. More …

Digeo: Discriminative geometry-aware learning for generalized few-shot object detection

J Ma, Y Niu, J Xu, S Huang, G Han… - Proceedings of the …, 2023 - openaccess.thecvf.com
Generalized few-shot object detection aims to achieve precise detection on both base
classes with abundant annotations and novel classes with limited training data. Existing …

A neural collapse perspective on feature evolution in graph neural networks

V Kothapalli, T Tirer, J Bruna - Advances in Neural …, 2024 - proceedings.neurips.cc
Graph neural networks (GNNs) have become increasingly popular for classification tasks on
graph-structured data. Yet, the interplay between graph topology and feature evolution in …

No fear of classifier biases: Neural collapse inspired federated learning with synthetic and fixed classifier

Z Li, X Shang, R He, T Lin… - Proceedings of the IEEE …, 2023 - openaccess.thecvf.com
Data heterogeneity is an inherent challenge that hinders the performance of federated
learning (FL). Recent studies have identified the biased classifiers of local models as the key …

Federated learning with bilateral curation for partially class-disjoint data

Z Fan, J Yao, B Han, Y Zhang… - Advances in Neural …, 2024 - proceedings.neurips.cc
Partially class-disjoint data (PCDD), a common yet under-explored data formation where
each client contributes a part of classes (instead of all classes) of samples, severely …

Proxymix: Proxy-based mixup training with label refinery for source-free domain adaptation

Y Ding, L Sheng, J Liang, A Zheng, R He - Neural Networks, 2023 - Elsevier
Due to privacy concerns and data transmission issues, Source-free Unsupervised Domain
Adaptation (SFDA) has gained popularity. It exploits pre-trained source models, rather than …