1
NOT_ADMIN 1 point ago +1 / -0

I work with many fortune 100 companies, and the US government. Many companies need to do object recognition of their conveyor lines, it needs to be generalized to many factories not just one. Even when there is real world data, often times the models need to be generalized and that’s very very hard to do with real world data. I’m talking about domain randomization because you don’t want to over fit a model. Often times, the client can provide some small amount of real world data, I then create a digital twin of what they need to recognize but because it’s virtual we can produce and annotate 100000 images with perfect segmentation masks, and fully randomized domains like lighting, orientation, noise etc so that the training data is extremely high quality and the chance for overfitting is minimal. Even for border security, being able to track migrants very far away; a simulated environment was created that looked similar to the land they were targeting but because it was virtual, we could randomize weather, lighting, trees, landscape water etc allowing for a much better and more flexible model.

1
NOT_ADMIN 1 point ago +1 / -0

Yes there is. In fact, many occasions, synthetic data is better. I work for an agency which creates Ai for fortune 100 companies and the US government. Maybe in 2016 synthetic data wasn't used...

0
NOT_ADMIN 0 points ago +1 / -1

I am a machine learning data generation developer…

0
NOT_ADMIN 0 points ago +1 / -1

You don't need real data. Synthetic data works, many times it's better.

34
NOT_ADMIN 34 points ago +34 / -0

Dear lord, we place our wants, desires, fears, anxieties, and hopes on you. We trust in you and know you love us all. Lift away these burdens, and keep us strong. Amen.

1
NOT_ADMIN 1 point ago +1 / -0

Quick question here: if the spider is as bad as they want us to think, why did they use a picture of someone willfully placing it on their hand? Or did they use a stock unrelated image for clickbait?

3
NOT_ADMIN 3 points ago +3 / -0

Anyone got a source location for this doc? one that can be used to redpill like a .gov link?

3
NOT_ADMIN 3 points ago +3 / -0

Thank you, I withdrawal my reservations.

5
NOT_ADMIN 5 points ago +5 / -0

Thank you, do you happen to know where to get the source document so that it can be verified that the words are less bold and were not manipulated before it got to you?

4
NOT_ADMIN 4 points ago +4 / -0

I don't understand the logic used to come up with what letter to highlight. Please elaborate on the objective pattern so that it is clear you aren't just looking for letters to make a message.

  • here is an example of why I want to see the objective pattern.
  • I used the same image, made up a word, and found it in the message HERE This example even gives a clearer pattern than yours which is a diagonal progression.
  • Keep in mind, "Dragon" is not a hidden message, it is simply a word I was looking for and found which is not how a secret message is found. Its an example of how not to find one.
  • This should not be stickied until there is a clear and objective pattern and once a pattern is established, the letter needs to be verified as rea with the same exact pattern.
view more: ‹ Prev Next ›