English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

With the ever-growing complexity of deep learning models for facerecognition, it becomes hard to deploy these systems in real life. Researchershave two options: 1) use smaller models; 2) compress their current models.Since the usage of smaller models might lead to concerning biases, compressiongains relevance. However, compressing might be also responsible for an increasein the bias of the final model. We investigate the overall performance, theperformance on each ethnicity subgroup and the racial bias of aState-of-the-Art quantization approach when used with synthetic and real data.This analysis provides a few more details on potential benefits of performingquantization with synthetic data, for instance, the reduction of biases on themajority of test scenarios. We tested five distinct architectures and threedifferent training datasets. The models were evaluated on a fourth datasetwhich was collected to infer and compare the performance of face recognitionmodels on different ethnicity.

Compressed Models Decompress Race Biases: What Quantized Models Forget for Fair Face Recognition