Digests » 124

this week's favorite

A penguin fish-recommender systems using multi-armed bandits

With the climate crisis raging on, I bet you are all thinking about the penguins. How are they going to find food with their ecosystem collapsing? It only makes sense that us humans take our responsibility to feed them. However, it would be preferable not to feed them manually, since Antarctica is a somewhat inhospitable place.

This AI Restores Old Photos with Damages Automatically!

A week ago Microsoft released a new paper called: Bringing Old Photos Back To Life. This is probably the best physical damage restoration for images I've ever seen. It also includes facial enhancement which is really accurate too.

12 Factors of reproducible Machine Learning in production

As a field, Machine Learning differs from traditional software development, but we can still borrow many learnings and adapt them to “our” industry. For the last few years, we’ve been doing Machine Learning projects in production, so beyond proof-of-concepts, and our goals where the same is in software development: reproducibility. So we built a pipeline orchestrator, strong automations and established a workflow to achieve exactly that.

How to Compare ROC Curves for Model Selection

The ROC curve was first developed in the early 1950s with the theory of electronic signal detection. One of the first applications was in the field of radar to detect weak signals in the presence of noise. The ROC methodology was later adapted and extensively used in epidemiology and medical research to reflect the relationship between sensitivity and specificity for a given test.

First Order Motion Model for Image Animation

Image animation consists of generating a video sequence so that an object in a source image is animated according to the motion of a driving video. Our framework addresses this problem without using any annotation or prior information about the specific object to animate.