nametube

It doesn't it repeats/generates pixels that minimize E, so the behaviour heavily depends on the choice of E. for inpainting E is the difference between their generated x and the true x but they mask out the error in the white areas. The model gets good results by just copying nearby pixels.

could you highlight which part this is ?. I think it can only fill in patterns that already exist in the image. I see a big artifact over the window but no lamp.

Evolutionary Algorithms are a subset of Population Based heuristics. I don't think its "souped up BS" to use the term.

People who aren't entrenched now have another service to choose from. people who are entrenched will be fought over on price and quality of service.

could you elaborate on that ?

I've never used Lyft but i think it works just like uber. this is what i found on their site

How Much Does a Lyft Ride Cost? Depends on where you’re headed and when. There is a base charge and a cost per mile and per minute, with extra added on when your pickup point is in Prime Time, which happens during higher-than-normal demand to encourage drivers to get on the road. For price estimates of rides in your city, check out the Lyft city page. Oh, and never worry about carrying cash — payment happens seamlessly through the app.

https://blog.lyft.com/posts/how-does-lyft-work

What kinds of models have you been playing around with ?

If you freeze the frame and look at the shell closely theres a trigger and stock that is warped. On the underside the fins have gun features as well.

You're going to have to back that up. This would be a clearly biased dataset.

could you elaborate on that?. They are the most popular languages for scientific computing.

This isn't a Microsoft product.

Something i've noticed recently is that my feed is flooded with the "x has become friends with y" type of notifications. I literally see 5-10 of them in a row, they make for a very monotonous feed experience. I think its supposed to be a nudge for me to add co-workers and acquaintances to facebook.

You might also be interested in openscoring https://github.com/openscoring/openscoring . It lets you run PMML models and score examples through a rest API.

from the article

> We developed ONNX together with Microsoft to bridge this gap and to empower AI developers to choose the framework that fits the current stage of their project and easily switch between frameworks as the project evolves.

ML workflows tend to be very heterogenous with people using everything from sklearn to DL4J to play with data. standardized serialization formats make it possible to bring results from those experiments back into a hardened productionized environment for serving the model.

While this is awesome I'm personally wondering when big companies like FB and MS are going to standardize on a format for predictive models in general. [0] PMML exists but there are very few companies that are standing behind it.

[0] http://dmg.org/pmml/v4-3/GeneralStructure.html

Image and Video codecs come under attack quite often see https://blog.sucuri.net/2016/05/imagemagick-remote-command-e...

In this context the image manipulation they do with pillow and the underlying libjpeg would be a potential source of vulnerabilities.

User generated content (especially images) are a great attack vector, what do you do to isolate/mitigate against attacks like that ?.

Maybe having the cards on the front page might help. You've invested a lot of screen realestate on explaining the product. You could probably make this https://www.placecard.me/try/ your homepage, have the instruction in the banner and have the product in front of the user right away.

Took a look at the site https://www.placecard.me/ and went through the flow of making a card and couldn't see a single place to pay for the template or order prints. Its possible I hit some kind of A/B test but i think the call to action might be buried too deep in the process.

I think the goal is to allow people to develop webapps with models built using neural network libraries like keras and tensorflow. This would greatly improve the distribution of applications that are powered by deep learning because you wont have to install a bunch of dependencies in order to use the app.

"running networks in the browser with TensorFire can be faster than running it natively with TensorFlow."

could you elaborate on this statement ?. What kinds of architectures does this hold true for ?.