Semi-Automatic Data Annotation guided by Feature Space Projection

TL;DR

This paper introduces a semi-automatic data annotation method that leverages feature space projection and semi-supervised learning to reduce manual labeling effort and improve classification accuracy, validated on MNIST and parasite images.

Contribution

The paper proposes a novel semi-automatic annotation approach combining feature space projection with semi-supervised learning, enhancing label propagation efficiency and accuracy.

Findings

Effective label propagation reduces manual annotation effort.

Improved classification accuracy on diverse datasets.

Visual analytics tools enhance human-machine collaboration.

Abstract

Data annotation using visual inspection (supervision) of each training sample can be laborious. Interactive solutions alleviate this by helping experts propagate labels from a few supervised samples to unlabeled ones based solely on the visual analysis of their feature space projection (with no further sample supervision). We present a semi-automatic data annotation approach based on suitable feature space projection and semi-supervised label estimation. We validate our method on the popular MNIST dataset and on images of human intestinal parasites with and without fecal impurities, a large and diverse dataset that makes classification very hard. We evaluate two approaches for semi-supervised learning from the latent and projection spaces, to choose the one that best reduces user annotation effort and also increases classification accuracy on unseen data. Our results demonstrate the added-value of visual analytics tools that combine complementary abilities of humans and machines for more effective machine learning.