Projects
Focused on ML research, typically with visual signals, some multi-modal. Experienced data scientist - data recognition and understanding, landmark detection, generative modeling, learning with minimal or imbalanced, and the byproduct of bias in ML with consideration to its impact on society, multi-modal databases (i.e., acquiring, designing, collecting, organizing, annotating, evaluating).
Families In the Wild Data and Benchmarks
Lead Researcher
Kinship recognition is challenging but comes with many applications. We have made significant progress these last ten years. We review the public resources and data challenges that enabled and inspired many to hone in on the views of automatic kinship recognition in the visual domain. The different tasks are described in technical terms and syntax consistent across the problem domain and the practical value.
Laplace Landmark Localization
Lead Researcher
Heatmaps generated by soft argmax-based models (middle block) and the proposed LaplaceKL (right block), each with heatmaps on the input images (left) and a zoomed-in view of an eye region (right). These heatmaps are confidence scores (i.e. probabilities) that a pixel is a landmark. softargmax-based methods generate highly scattered mappings (low certainty), while the same network trained with our loss is concentrated (i.e. high certainty).
Balanced Faces In the WIld
Lead Researcher
This project investigates bias in automatic facial recognition (FR). Specifically, subjects are grouped into predefined subgroups based on gender, ethnicity, and soon-to-be age. For this, we propose a novel image collection called Balanced Faces in the Wild (BFW), which is balanced across eight subgroups (i.e., 800 face images of 100 subjects, each with 25 face samples).
Dissertation
Author
Joseph P Robinson | PhD Dissertation, 2020 | Northeastern University
2021
-
Joseph P Robinson et al. "Balancing Biases and Preserving Privacy on Balanced Faces in the Wild." arXiv preprint arXiv:2103.09118 (2021). [paper]
-
Joseph P Robinson et al. "Families In Wild Multimedia (FIW-MM): A Multi-Modal Database for Recognizing Kinship." in IEEE Transaction on Multimedia (2021). [paper]
-
Joseph P Robinson et al. "Survey on the Analysis and Modeling of Visual Kinship: A Decade in the Making." in IEEE Trans. on Pattern Analysis & Machine Intelligence (2021). [paper]
2020
-
Joseph P Robinson et al. "Face Recognition: Too Bias, or Not Too Bias?." in CVPR Workshop (2020). [paper]​
-
Joseph P Robinson et al. "Recognizing Families In the Wild (RFIW): The 4th Edition." in IEEE Automatic Face and Gesture Recognition (2020). [paper]​​
-
Yu Yin, Joseph P Robinson, Yulun Zhang, Yun Fu. "Joint Super-Resolution and Alignment of Tiny Faces." in AAAI (2020). [paper]
2019
2018
2017
-
Joseph P Robinson et al. "Recognizing Families In the Wild (RFIW): Data Challenge Workshop," ACM MM: Workshop on RFIW (2017). [paper, proceedings, contents]
2016
-
Joseph P Robinson et al. “Families In the Wild (FIW): large-scale kinship image database and benchmarks." In Proc. of the ACM on Multimedia Conference (2016). [paper]