Overview

The ICCV 2025 Workshop on Curated Data for Efficient Learning (CDEL) seeks to advance the understanding and development of data-centric techniques that improve the efficiency of training large-scale machine learning models. As model sizes continue to grow and data requirements scale accordingly, this workshop brings attention to the increasingly critical role of data quality, selection, and synthesis in achieving high model performance with reduced computational cost. Rather than focusing on ever-larger datasets and models, CDEL emphasizes the curation and distillation of high-value data—leveraging techniques such as dataset distillation, data pruning, synthetic data generation, and sampling optimization. These approaches aim to reduce redundancy, improve generalization, and enable learning in data-scarce regimes. The workshop will bring together researchers and practitioners from vision, language, and multimodal learning to share insights and foster collaborations around efficient, scalable, and sustainable data-driven machine learning.

Invited Speakers

Antonio Torralba
Massachusetts Institute of Technology

Alyosha Efros
University of California, Berkeley

Olga Russakovsky
Princeton University

Sara Beery
Massachusetts Institute of Technology

Zhuang Liu
Princeton University

Schedule

Date: October 18, 2025
Time: 8:30 AM – 5:30 PM
Location: ICCV Workshop Venue (TBA)

08:30 – 08:55	Welcome & Coffee
09:00 – 09:35	Invited Talk: Antonio Torralba
09:35 – 10:10	Invited Talk: Alyosha Efros
10:10 – 10:45	Oral Presentations I
10:50 – 12:00	Poster Session
12:00 – 13:25	Lunch Break
13:30 – 14:25	Invited Talk: Olga Russakovsky
14:30 – 15:05	Invited Talk: Sara Beery
15:05 – 15:40	Invited Talk: Zhuang Liu
15:40 – 16:25	Oral Presentations II
16:30 – 17:00	Panel Discussion & Closing

Call for Papers

We welcome submissions on all topics related to the curation of training data.
Some potential topics include:

Data Pruning: How can we eliminate redundant or low-quality samples from large datasets?
Synthetic Data: How can we use generative models to create or augment datasets?
Dataset Distillation: How can we learn tiny datasets of highly-efficient synthetic samples?
Obscure Domains: How can we train models in areas where existing data is extremely scarce?
Future Directions: What problems in data-centric AI can we expect in the near future?

Submission Details:
We accept submissions of both long conference-style papers (8 pages) and short extended abstracts (4 pages). Authors of accepted long papers have the option of having their work published in the ICCV workshop proceedings if they do not violate dual-submission guidelines.

We also welcome submissions of work currently in submission or recently accepted to other venues, but these will not be published in the workshop proceedings (but may still be presented at our workshop).

Please sign up here if you’d like to volunteer as a reviewer.