paper-with-me

Papers

Data-Efficient Learning via Minimizing Hyperspherical Energy

2022-06-30 · Xiaofeng Cao, Weiyang Liu, Ivor W. Tsang

Deep learning on large-scale data is dominant nowadays. The unprecedented scale of data has been arguably one of the most important driving forces for the success of deep learning. However, there still exist scenarios where collecting data or labels could be extremely expensive, e.g., medical imaging and robotics. To fill up this gap, this paper considers the problem of data-efficient learning from scratch using a small amount of representative data. First, we characterize this problem by active learning on homeomorphic tubes of spherical manifolds. This naturally generates feasible hypothesis class. With homologous topological properties, we identify an important connection -- finding tube manifolds is equivalent to minimizing hyperspherical energy (MHE) in physical geometry. Inspired by this connection, we propose a MHE-based active learning (MHEAL) algorithm, and provide comprehensive theoretical guarantees for MHEAL, covering convergence and generalization analysis. Finally, we demonstrate the empirical performance of MHEAL in a wide range of applications on data-efficient learning, including deep clustering, distribution matching, version space sampling and deep active learning.

📄 PDF Abstract BibTeX arXiv:2206.15204

Code (0)

등록된 구현이 없습니다.

Tasks

Active LearningDeep Clustering

Similar Papers 제목 키워드 기반

Regularizing Neural Networks via Minimizing Hyperspherical Energy

2019-06-12 · CVPR 2020 6 · Rongmei Lin, Weiyang Liu, Zhen Liu, Chen Feng 외

Inspired by the Thomson problem in physics where the distribution of multiple propelling electrons on a unit sphere can be modeled via minimizing some potential energy, hyperspherical energy minimization has demonstrated…

Eccentric Regularization: Minimizing Hyperspherical Energy without explicit projection

2021-04-23 · Xuefeng Li, Alan Blair

Several regularization methods have recently been introduced which force the latent activations of an autoencoder or deep neural network to conform to either a Gaussian or hyperspherical distribution, or to minimize the …

DiversityImage GenerationRepresentation Learning

Parameter-efficient Fine-tuning in Hyperspherical Space for Open-vocabulary Semantic Segmentation

2024-05-29 · CVPR 2025 1 · Zelin Peng, Zhengqin Xu, Zhilin Zeng, Yaoming Wang 외

Open-vocabulary semantic segmentation seeks to label each pixel in an image with arbitrary text descriptions. Vision-language foundation models, especially CLIP, have recently emerged as powerful tools for acquiring open…

Open Vocabulary Semantic SegmentationOpen-Vocabulary Semantic Segmentationparameter-efficient fine-tuningSemantic Segmentation

Bridging Mini-Batch and Asymptotic Analysis in Contrastive Learning: From InfoNCE to Kernel-Based Losses

2024-05-28 · Panagiotis Koromilas, Giorgos Bouritsas, Theodoros Giannakopoulos, Mihalis Nicolaou 외

What do different contrastive learning (CL) losses actually optimize for? Although multiple CL methods have demonstrated remarkable representation learning capabilities, the differences in their inner workings remain lar…

Contrastive LearningRepresentation Learning

Hyperspherically Regularized Networks for Self-Supervision

2021-04-29 · Aiden Durrant, Georgios Leontidis

Bootstrap Your Own Latent (BYOL) introduced an approach to self-supervised learning avoiding the contrastive paradigm and subsequently removing the computational burden of negative sampling associated with such methods. …

DiversitySelf-Supervised Learning