Joel Brogan

Dr. Joel Brogan

orcid

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 March 2018 in arXiv
Verified
Preprint
Open access
PDF file
SBSandipan Banerjee
Sandipan Banerjee
JBJoel Brogan
Joel Brogan
Janez KrižajJanez Križaj
Janez Križaj
ABAparna Bharati
Aparna Bharati
BRBrandon RichardWebster
Brandon RichardWebster
WSWalter Scheirer
Walter Scheirer
Face recognition performance has improved remarkably in the last decade. Much of this success can be attributed to the development of deep learning techniques such as convolutional neural networks (CNNs). While CNNs have pushed the state-of-the-art forward, their training process requires a large amount of clean and correctly labelled training data. If a CNN is intended to tolerate facial pose, then we face an important question: should this training data be diverse in its pose distribution, or should face images be normalized to a single pose in a pre-processing step? To address this question, we evaluate a number of popular facial landmarking and pose correction algorithms to understand their effect on facial recognition performance. Additionally, we introduce a new, automatic, single-image frontalization scheme that exceeds the performance of current algorithms. CNNs trained using sets of different pre-processing methods are used to extract features from the Point and Shoot Challenge (PaSC) and CMU Multi-PIE datasets. We assert that the subsequent verification and recognition performance serves to quantify the effectiveness of each pose correction scheme.

2 December 2019 in arXiv
Verified
Preprint
Open access
PDF file
JBJoel Brogan
Joel Brogan
ABAparna Bharati
Aparna Bharati
DMDaniel Moreira
Daniel Moreira
KBKevin Bowyer
Kevin Bowyer
PFPatrick Flynn
Patrick Flynn
WSWalter Scheirer
Walter Scheirer
Images from social media can reflect diverse viewpoints, heated arguments, and expressions of creativity, adding new complexity to retrieval tasks. Researchers working onContent-Based Image Retrieval (CBIR) have traditionally tuned their algorithms to match filtered results with user search intent. However, we are now bombarded with composite images of unknown origin, authenticity, and even meaning. With such uncertainty, users may not have an initial idea of what the results of a search query should look like. For instance, hidden people, spliced objects, and subtly altered scenes can be difficult for a user to detect initially in a meme image, but may contribute significantly to its composition. We propose a new approach for spatial verification that aims at modeling object-level regions dynamically clustering keypoints in a 2D Hough space, which are then used to accurately weight small contributing objects within the results, without the need for costly object detection steps. We call this method Objects in Scene to Objects in Scene (OS2OS) score, and it is optimized for fast matrix operations on CPUs. OS2OS performs comparably to state-of-the-art methods in classic CBIR problems, on the Oxford5K, Paris 6K, and Google-Landmarks datasets, without the need for bounding boxes. It also succeeds in emerging retrieval tasks such as image composite matching in the NIST MFC2018 dataset and meme-style composite imagery fromReddit.