ACTIVELOOP AT

CVPR

Activeloop, the dataset optimization company,
is excited to be a part of CVPR 2021!


For this edition of CVPR, Activeloop is proud to host two panels on the future of data and machine learning infrastructure for Computer Vision applications.

Join the conversation with leading experts from Google, Microsoft, RISELab at UC Berkeley, Anyscale, HuggingFace, Weights and Biases and more!

The Future of Datasets

Databases, data lakes, and data warehouses are unfit for unstructured data types such as images, videos, and text. Data 2.0 allows storing computer vision data, including metadata, in multi-dimensional arrays, locally or on the cloud.

[removal.ai]_tmp-60b92da6bc3cb.png

CVPR pre-game:
The Future of Datasets

Friday, June 18th at 12 pm EDT / 9 am PST

Screen Shot 2021-06-03 at 10.33.45 PM.pn

Jeff
Boudier

[removal.ai]_tmp-60b92da6bc3cb.png

Head of Growth and Product at HuggingFace

Screen Shot 2021-06-03 at 10.35.07 PM.pn

Richard
Socher

Screenshot%202021-06-09%20at%200.36_edit

CEO at

You.com
Co-Author, ImageNet

Screen Shot 2021-06-10 at 2.23.31 PM.png

Olga Russakovsky

[removal.ai]_tmp-60c25bc62f23f (1).png

Assistant Professor at Princeton

Co-lead Author of ImageNet Challenge

Screen Shot 2021-06-04 at 2.48.46 PM.png

Jianing
Wei

[removal.ai]_tmp-60b92d4522430.png

Tech Lead Manager at Google

Screen Shot 2021-06-03 at 10.38.21 PM.pn

Joseph Gonzalez

[removal.ai]_tmp-60bfb6ea5e193.png

UC Berkeley
RISE Lab

Sid_podcast_1400x788_No-logos-240x180.pn

Siddhartha
Sen

microsoft-logo-hd-11563140791lh7zg8bhmr-

Principal Researcher at

Microsoft

An automated, end-to-end ML pipeline would accelerate research. Far too often, however, we concentrate on optimizing models rather than optimizing data. In this discussion, we discuss ways to get the most out of datasets.

 
 

CVPR Panel: Next-Gen ML Infrastructure For Computer Vision

Monday, June 21st at 3 pm EDT / 12 pm PST

Screen Shot 2021-06-03 at 10.40.05 PM.pn

Dillon
Erb

5dd6613ca930f930ebbdcd4b_PaperspaceCircl

CEO at

Paperspace

Screen Shot 2021-06-03 at 10.41.27 PM.pn

Tianqi
Chen

214d154e-e318-4d78-a9e8-eddcc529e046_thu

CTO at 

OctoML

Screen Shot 2021-06-03 at 10.44.25 PM.pn

Glenn
Jocher

[removal.ai]_tmp-60b9319db1d51.png

CEO at Ultralytics

Creator of YOLOv5

Screen Shot 2021-06-04 at 5.25.28 PM.png

Davit Buniatyan

light-cc59432478ad93fe28fc2d716e99ef01.p

CEO at

Activeloop

Screen Shot 2021-06-03 at 10.45.41 PM.pn

Lukas
Biewald

1551466533202-removebg-preview.png

CEO at

Weights & Biases

Screen Shot 2021-06-03 at 10.37.18 PM.pn

Waleed Kadous

ANSC_BM_social_icon_01.png

Head of Engineering at Anyscale

Screen Shot 2021-06-03 at 10.42.45 PM.pn

Josh
Tobin

gantry-logo-1600px.png

CEO at

Gantry

Screen Shot 2021-06-08 at 2.33.42 PM.png
1565008950053.png

CEO & Co-Founder at D2iQ

Previously, we explored a data-centric approach to ML training. Unfortunately there aren't many existing solutions for dataset management and optimization. In this discussion, we explore tooling and infrastructure to get the most out of data.

 

Drop in to our CVPR booth. 

ABOUT ACTIVELOOP

Activeloop, a dataset optimization company, seamlessly manages data for deep learning applications. Activeloop automatically connects unstructured data to machine learning models. Our open-source package Hub (3.3K+ stars) enables data streaming, scalable machine learning pipelines, and dataset version control for distributed workloads.

Activeloop’s control panel for deep learning allows companies to easily access, visualize and improve their datasets to build great models. Activeloop’s stack is used by teams at Google, Waymo, and Red Cross.