The least-squares method for classification is based on linearly separating 2 or more classes. In this article, I’m going to show you how to create a Python program to classify images with digits from 0–9 using only NumPy and PIL.
Continue reading “Using least squares-based classification to detect digits”
Sometimes you might want to change part of your lenses and replace them with an image or object or even an effect. Segmentation allows you to do that while using segmentation textures.
For instance, you can change the background by adding the texture you want or changing the user’s hair by adding color or a texture to it.
Types of segmentation textures include portrait background, portrait hair, portrait shoulder, portrait face, portrait head, sky, and body. We will look at what they do later on in this article.
So you came here—let me guess—it’s either you’re in a data science competition and you read somewhere about how winners of most competitions win with ensembles, or you’re just a curious data scientist who wants to learn about ensembles.
Either way, understanding how ensembles work is a very important knowledge and as data scientists and machine learning engineers, you should be able to employ the skills behind them.
Research has shown that a majority of the time, ensembles will outperform a single model, and it’s the recommended technique for maximizing accuracy or reducing errors in a machine learning model.
In this tutorial, we’ll cover how to implement the neural-style algorithm that’s based on this paper.
What is neural style transfer?Neural style transfer is a technique used to generate images in the style of another image. The neural-style algorithm takes a content-image as input, a style image, and returns the content image as if it were painted using the artistic style of the style image.
LUT or, Look Up Table, is a photo or image filter that enhances and changes the color tone and grading of your image. It essentially can convert colors and details in a source file to a new destination state.
Using a custom lookup table or “LUT” allows us to have free range on creating our own custom color corrections through external third party programs and bring them directly into Lens Studio.
Machine learning (ML) began its ascent into the medical industry when it acquired the ability to detect visual patterns between images—a skill doctors and technicians take years to master.
Specifically, ML models for computer vision tasks in the medical field train on datasets of separated images to learn to recognize their similarities and differences.
Continue reading “Detecting Skin Cancer on iOS with Xcode and Create ML”
Making lenses is a difficult endeavor, but for Shawn there’s another layer: he’s colorblind. The young creator has already made a number of lenses that allow users to see the world the way he does.
On top of a solar system themed series of lenses, he’s created a number of fun, subtle filters. I chatted with the creator to learn more.
If you got past the jargon of the title, you probably have at least a passing interest in computer vision. However, fear not! This is going to be a fairly gentle walk-through of some of my projects at the intersection of Machine Learning and Augmented Reality.
They all share a common denominator: feature point extraction. InstaSaber, Say BARK!, and the puppet videos you see below are a few examples.
Continue reading “Real-Time 2D/3D Feature Point Extraction from a Mobile Camera”
Throughout the Android chat app development series, we’ve built an app that allows users to register, send/receive text messages, open conversations, and receive notifications for incoming messages.
One additional essential feature for chat apps is the ability to update the active, live conversation. Once a new message arrives for the currently active conversation, it should be updated to reflect the new changes (either sent/received messages).
Continue reading “Live Conversation Updates for New Messages in an Android Chat Application”
In the past few years we’ve seen many startups and even mature companies coming up with new mobile apps or features powered by machine learning and AI. These features require some heavy, real-time processing by neural networks.
The potential killers of these ML-powered experiences? Data roundtrips for inference, the cost of backend servers to support millions of devices, concerns surrounding user data privacy.
Continue reading “Comparing Mobile Machine Learning Frameworks”
