The Ground Truth #10

The Ground Truth #10

2022 was a year full of big technology. In this newsletter, we review our most popular content and our biggest product releases.?Scroll down to read more…

From the Industry

In Time Magazine’s?Best Inventions of 2022, the presence of big developments in AI was no surprise to anyone. Our predictions were right on target when it came to Dall-E 2 - it’s a game changer. But, for computer vision engineers in particular, it was the inclusion of?Instant NeRF?from NVIDIA that is most notable. NVIDIA took a process that “required specialized equipment, expertise and lots of time and money” and changed that to a few minutes and a few pictures. The result: Realistically rendered 3D scenes from static images or videos.

2023 should also be groundbreaking and we’re looking forward to it.

Top Blog & Paper Review

Paper Review

SynFace: Face Recognition with Synthetic Data?

This paper answers the question: What is the optimal way to incorporate synthetically generated images into the training of face recognition models? To address this question and more, the authors propose?SynFace, an approach that combines real and synthetic facial data to improve the performance of face recognition models. Read the review?here.?

Blog

?6 Papers to Watch at CVPR 2022

CVPR always has great reveals and we reviewed 6 papers to watch including 2 new types of GANs, Task2Sim, removing glasses and shadows with synthetic data, Mip-NeRF 360 and more. Read the blog?here.?

Top Product Updates

Humans in Context

We were excited to expand our humans in motion synthetic data this year. Across a variety of domains - in-cabin automotive, smart fitness, smart office, and home security - you can get the data you need. You can choose from a variety of identities and hundreds of behaviors in multiple environments with different camera and lighting settings.?

API & SDK

This year, we announced our API & SDK that allows customers to use code to generate their synthetic data via Python. With the API, all our 50K identities can be controlled granularly, down to the smallest detail.??

If you haven’t tried our API & SDK or worked with our humans in motion, you can now with our free trial. Start?here.?

No alt text provided for this image
Data for 3D reconstruction generated by the Datagen API

Top Content

Designing a Synthetic Data Solution




eBook

Designing a Synthetic Data Solution

Read?here.




No alt text provided for this image



Benchmark

Facial Landmark Detection Using Synthetic Data

Read?here.?



Michael J. Black Podcast


Podcast

Body Models: Driving the Age of the Avatar

with Michael J. Black?

Listen now.

要查看或添加评论,请登录

Datagen的更多文章

社区洞察

其他会员也浏览了