Evaluation of Face Datasets as Tools for Assessing the Performance of Face Recognition Methods.

Int J Comput Vis

Laboratory of Genetics, National Institute on Aging, National Institutes of Health, 333 Cassell Dr., Baltimore, MD 21224, Tel: 410-558-8682 , Email:

Published: January 2008

Face datasets are considered a primary tool for evaluating the efficacy of face recognition methods. Here we show that in many of the commonly used face datasets, face images can be recognized accurately at a rate significantly higher than random even when no face, hair or clothes features appear in the image. The experiments were done by cutting a small background area from each face image, so that each face dataset provided a new image dataset which included only seemingly blank images. Then, an image classification method was used in order to check the classification accuracy. Experimental results show that the classification accuracy ranged between 13.5% (color FERET) to 99% (YaleB). These results indicate that the performance of face recognition methods measured using face image datasets may be biased. Compilable source code used for this experiment is freely available for download via the internet.

Download full-text PDF

Source
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2529479PMC
http://dx.doi.org/10.1007/s11263-008-0143-7DOI Listing

Publication Analysis

Top Keywords

face datasets
12
face recognition
12
recognition methods
12
face
10
performance face
8
face image
8
classification accuracy
8
image
5
evaluation face
4
datasets
4

Similar Publications

Motivation: Ensuring connectivity and preventing fractures in tubular object segmentation are critical for downstream analyses. Despite advancements in deep neural networks (DNNs) that have significantly improved tubular object segmentation, existing methods still face limitations. They often rely heavily on precise annotations, hindering their scalability to large-scale unlabeled image datasets.

View Article and Find Full Text PDF

Patients with inflammatory bowel disease (IBD) face an elevated risk of developing colorectal cancer (CRC). Endoscopic surveillance is a cornerstone in CRC prevention, enabling early detection and intervention. However, despite recent advancements, challenges persist.

View Article and Find Full Text PDF

This case study highlights the use of cinematic rendering (CR) in preoperative planning for the excision of a cyst in the oral and maxillofacial region of a 60-year-old man. The patient presented with a firm, non-tender mass in the right cheek, clinically suspected to be an epidermoid cyst. Conventional imaging, including dental magnetic resonance imaging (MRI) protocols, confirmed the lesion's size, location, and benign nature.

View Article and Find Full Text PDF

AutoGP: An Intelligent Breeding Platform for Enhancing Maize Genomic Selection.

Plant Commun

January 2025

National Key Laboratory of Crop Genetic Improvement, Huazhong Agricultural University, Wuhan 430070, China; Hubei Hongshan Laboratory, Hubei, China. Electronic address:

In the face of climate change and the growing global population, there is an urgent need to accelerate the development of high-yielding crop varieties. To this end, vast amounts of genotype-to-phenotype data have been collected, and many machine learning (ML) models have been developed to predict phenotype from a given genotype. However, the requirement for high densities of single-nucleotide polymorphisms (SNPs) and the labor-intensive collection of phenotypic data are hampering the use of these models to advance breeding.

View Article and Find Full Text PDF

Translating biological knowledge from Arabidopsis to crop species is important to advance agriculture and secure food production in the face of dwindling fertilizer resources and biotic and abiotic stresses. However, it is often not trivial to identify functional homologs (orthologs) of Arabidopsis genes in crops. Combining sequence and expression data can improve the correct prediction of orthologs.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!