PeopleofSuperintelligence

Researcher

Priya Goyal

Priya Goyal is a computer vision researcher whose work spans object detection, self-supervised learning, fairness, and agentic systems. She earned bachelor's and master's degrees in mathematics and scientific computing at IIT Kanpur in 2015, then was accepted into Facebook AI Residency in 2016. Her research includes Focal Loss for object detection, a separate method for training ImageNet in one hour, SwAV, VISSL, and SEER. In 2022, she coauthored fairness research for computer vision models. She joined DatologyAI as a founding member in January 2024; as of October 2026, she is building agentic systems and evaluation methods focused on data and feedback.

Works
13
Citations
46,583
h-index
12
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2015

    Mathematics at IIT Kanpur

    She earns bachelor's and master's degrees in mathematics and scientific computing at IIT Kanpur in 2015.

  2. 2016

    Facebook AI Residency

    In spring 2016, she is accepted into Facebook AI Residency, a program designed to train software engineers to become AI researchers.

  3. 2017

    Focal loss and ImageNet training

    At ICCV 2017, she coauthors Focal Loss for Dense Object Detection, which down-weights the loss on well-classified examples to address foreground-background imbalance. In a separate 2017 study, she coauthors a Caffe2 system that trains ResNet-50 with minibatches of 8,192 images on 256 GPUs in one hour while matching small-minibatch accuracy.

  4. 2020

    SwAV visual representations

    In 2020, she coauthors SwAV, an online algorithm that clusters images and keeps cluster assignments consistent across different views of each image. The paper reports 75.3 percent top-1 ImageNet accuracy with ResNet-50.

  5. 2021

    VISSL, SEER, and FSDP

    In 2021, she coauthors VISSL, a FAIR library for self-supervised image learning, and work on SEER, whose 1.3-billion-parameter RegNetY reaches 84.2 percent top-1 ImageNet accuracy after training on one billion random images with 512 GPUs. In a 2021 Facebook AI Research Engineering post, she coauthors the description of Fully Sharded Data Parallel, which shards model parameters across workers and supports training larger models with fewer GPUs.

  6. 2022

    Fairness indicators for vision

    In 2022, she coauthors Fairness Indicators for Systematic Assessments of Visual Feature Extractors, which proposes three indicators to quantify harms and biases in computer vision models. The paper sets out protocols for applying the indicators across a range of models.

  7. January 2024

    Joining DatologyAI

    After work at FAIR and Google DeepMind, she joins DatologyAI as a founding member in January 2024. Her work there covers data quality and evaluation for AI training.

  8. As of October 2026

    Building agentic systems

    As of October 2026, she is an AI research and engineering leader building agentic systems. Her current focus includes agent evaluations, synthetic data, data quality, and feedback that helps agents improve.

Career

  1. From 2024DatologyAIFounding Member
  2. 2017–2022Meta

Education

  1. 2015Indian Institute of Technology KanpurBachelor's and master's degrees

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. Focal Loss for Dense Object Detection2017 · IEEE International Conference on Computer Vision (ICCV) · 30,170 citationsWith Kaiming He, Piotr Dollár, Ross Girshick and Tsung-Yi Lin
  2. Unsupervised Learning of Visual Features by Contrasting Cluster Assignments2020 · 2,499 citationsWith Julien Mairal and Piotr Bojanowski
  3. Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour2017 · 2,293 citationsWith Kaiming He, Piotr Dollár, Ross Girshick and Yangqing Jia
  4. Scaling and Benchmarking Self-Supervised Visual Representation Learning2019 · IEEE/CVF International Conference on Computer Vision (ICCV) · 341 citationsWith Abhinav Gupta
  5. Tensor Comprehensions: Framework-Agnostic High-Performance Machine Learning Abstractions2018 · 250 citations

From OpenAlex, as of 21 October 2026.

  1. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  2. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  3. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes Google. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.
  4. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Google. We estimate that alumni of 17 major AI labs went on to found or lead 350 other organizations, including 226 they founded, and we trace 472 career moves directly between 23 labs since 2010.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Priya Goyal yet. Send us a free-licensed photo.

  1. You Don't Need a Masters/PhD – How These 9 Engineers Broke Into ML trybackprop.com. Checked 20 October 2026.
  2. Focal Loss for Dense Object Detection openaccess.thecvf.com. Checked 20 October 2026.
  3. Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour arxiv.org. Checked 20 October 2026.
  4. Unsupervised Learning of Visual Features by Contrasting Cluster Assignments arxiv.org. Checked 20 October 2026.
  5. Self-supervised Pretraining of Visual Features in the Wild arxiv.org. Checked 20 October 2026.
  6. Fully Sharded Data Parallel: faster AI training with fewer GPUs engineering.fb.com. Checked 20 October 2026.
  7. README.md - facebookresearch/vissl github.com. Checked 20 October 2026.
  8. Fairness Indicators for Systematic Assessments of Visual Feature Extractors arxiv.org. Checked 20 October 2026.
  9. Priya Goyal prigoyal.github.io. Checked 20 October 2026.