CVJul 31, 2018

The Devil of Face Recognition is in the Noise

arXiv:1807.11649v1213 citationsHas Code
Originality Incremental advance
AI Analysis

This work addresses data quality issues for researchers and practitioners in face recognition, though it is incremental as it builds on existing datasets and methods.

The paper tackles the problem of label noise in face recognition datasets by analyzing its impact on model accuracy and showing that many more samples are needed with noisy data to achieve the same accuracy as with clean data. It contributes cleaned subsets of existing datasets and a new noise-controlled dataset, IMDb-Face, to facilitate research.

The growing scale of face recognition datasets empowers us to train strong convolutional networks for face recognition. While a variety of architectures and loss functions have been devised, we still have a limited understanding of the source and consequence of label noise inherent in existing datasets. We make the following contributions: 1) We contribute cleaned subsets of popular face databases, i.e., MegaFace and MS-Celeb-1M datasets, and build a new large-scale noise-controlled IMDb-Face dataset. 2) With the original datasets and cleaned subsets, we profile and analyze label noise properties of MegaFace and MS-Celeb-1M. We show that a few orders more samples are needed to achieve the same accuracy yielded by a clean subset. 3) We study the association between different types of noise, i.e., label flips and outliers, with the accuracy of face recognition models. 4) We investigate ways to improve data cleanliness, including a comprehensive user study on the influence of data labeling strategies to annotation accuracy. The IMDb-Face dataset has been released on https://github.com/fwang91/IMDb-Face.

Code Implementations2 repos
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes