Collaborative Unlabeled Data Optimization

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Shang, Xinyi, Sun, Peng, Liu, Fengyuan, Lin, Tao
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918158111080448
author Shang, Xinyi
Sun, Peng
Liu, Fengyuan
Lin, Tao
author_facet Shang, Xinyi
Sun, Peng
Liu, Fengyuan
Lin, Tao
contents This paper pioneers a novel data-centric paradigm to maximize the utility of unlabeled data, tackling a critical question: How can we enhance the efficiency and sustainability of deep learning training by optimizing the data itself? We begin by identifying three key limitations in existing model-centric approaches, all rooted in a shared bottleneck: knowledge extracted from data is locked to model parameters, hindering its reusability and scalability. To this end, we propose CoOpt, a highly efficient, parallelized framework for collaborative unlabeled data optimization, thereby effectively encoding knowledge into the data itself. By distributing unlabeled data and leveraging publicly available task-agnostic models, CoOpt facilitates scalable, reusable, and sustainable training pipelines. Extensive experiments across diverse datasets and architectures demonstrate its efficacy and efficiency, achieving 13.6% and 6.8% improvements on Tiny-ImageNet and ImageNet-1K, respectively, with training speedups of $1.94 \times $ and $1.2 \times$.
format Preprint
id arxiv_https___arxiv_org_abs_2505_14117
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Collaborative Unlabeled Data Optimization
Shang, Xinyi
Sun, Peng
Liu, Fengyuan
Lin, Tao
Machine Learning
Artificial Intelligence
This paper pioneers a novel data-centric paradigm to maximize the utility of unlabeled data, tackling a critical question: How can we enhance the efficiency and sustainability of deep learning training by optimizing the data itself? We begin by identifying three key limitations in existing model-centric approaches, all rooted in a shared bottleneck: knowledge extracted from data is locked to model parameters, hindering its reusability and scalability. To this end, we propose CoOpt, a highly efficient, parallelized framework for collaborative unlabeled data optimization, thereby effectively encoding knowledge into the data itself. By distributing unlabeled data and leveraging publicly available task-agnostic models, CoOpt facilitates scalable, reusable, and sustainable training pipelines. Extensive experiments across diverse datasets and architectures demonstrate its efficacy and efficiency, achieving 13.6% and 6.8% improvements on Tiny-ImageNet and ImageNet-1K, respectively, with training speedups of $1.94 \times $ and $1.2 \times$.
title Collaborative Unlabeled Data Optimization
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2505.14117