Data-Efficient Challenges in Visual Inductive Priors: A Retrospective

TL;DR

This paper reviews challenges in training deep learning models for computer vision with limited data, highlighting the role of inductive priors, data augmentation, and ensemble methods in improving data efficiency.

Contribution

It organizes and analyzes the VIPriors workshops, showcasing novel approaches that incorporate priors and data augmentation to enhance data-efficient deep learning.

Findings

Ensemble methods combining Transformers and CNNs improve performance.

Heavy data augmentation is crucial for data-efficient training.

Prior knowledge-based methods contribute to success in limited-data scenarios.

Abstract

Deep Learning requires large amounts of data to train models that work well. In data-deficient settings, performance can be degraded. We investigate which Deep Learning methods benefit training models in a data-deficient setting, by organizing the "VIPriors: Visual Inductive Priors for Data-Efficient Deep Learning" workshop series, featuring four editions of data-impaired challenges. These challenges address the problem of training deep learning models for computer vision tasks with limited data. Participants are limited to training models from scratch using a low number of training samples and are not allowed to use any form of transfer learning. We aim to stimulate the development of novel approaches that incorporate prior knowledge to improve the data efficiency of deep learning models. Successful challenge entries make use of large model ensembles that mix Transformers and CNNs, as well as heavy data augmentation. Novel prior knowledge-based methods contribute to success in some entries.