Seth Hendrickson

Software engineer @stripe

[Seth Hendrickson @shendrickson16]

@vykthur @MSFTResearch 👏👏👏

[Lou of the Candy Corn @SynStalker]

#code Our Manifesto

[François Chollet @fchollet]

I would rather press [▲] 50 times than actually type a command in the terminal

[Joe Biden @JoeBiden]

Wear a mask.

[Seth Hendrickson @shendrickson16]

@_cjwallace @cloudera @FastForwardLabs Congrats!

[Seth Hendrickson @shendrickson16]

@b0rk How Operating Systems work!

[Pranay Pathole @PPathole]

Effective debugging:

print("got this far")
...

print("got this far1")
...

...

...

...

print("got this far7")

[grant @GrantCuster]

We @fastforwardlabs recently released our sentiment analysis prototype Textflix. It uses movie reviews to show how natural language processing can unlock the data embedded in large amounts of unstructured text: https://t.co/MHBaE5ut88
#MachineLearning #NLP #dataviz
Design thread:

[Cloudera Fast Forward Labs @FastForwardLabs]

How can weak supervision help when you have limited labeled data? In our latest blogpost, we explore how to programmatically create labels with weak supervision. https://t.co/yUXhUjZBZe

[Seth Hendrickson @shendrickson16]

The visualization needs work 😆 @GrantCuster

[Seth Hendrickson @shendrickson16]

#DeepLearning fun with tennis and computer vision. Automatically decode raw video to player-tracking data. This is works from a labeled dataset of only about 100 frames. Posting a few examples.

[Seth Hendrickson @shendrickson16]

This just uses a single, non-customized Faster-RCNN (+keypoints) network. Originally there were separate stages for finding the court, segmenting point/non-point frames, & detecting players. But a single FRCNN network can do all this at once. A little surprised how well it works.

[Seth Hendrickson @shendrickson16]

@davidtalby @hmason @FastForwardLabs No unfortunately not.

[Seth Hendrickson @shendrickson16]

@davidtalby @hmason @FastForwardLabs Hey David, could we get in touch via direct message or e-mail? I'm shendrickson at cloudera dot com

[Cloudera Fast Forward Labs @FastForwardLabs]

Our favorite research, talks, art and other stuff from 2018 https://t.co/HFJX0DAqQl

[Seth Hendrickson @shendrickson16]

I wrote a thing about transfer learning, which has cat pictures and hand drawn robots... so, yeah, can't really think of a more convincing reason to read it than that. https://t.co/i2sgP6PkjW

[Cloudera Fast Forward Labs @FastForwardLabs]

We have a new post up on our blog! Check out @shendrickson16's take on #DeepLearning made easier with transfer learning. https://t.co/tISa6T2oFC