Blog

We’re living in an era of transition from inventions of the Renaissance to the technological revolution of automation. Even so, a lot of valuable records and information are still locked inside unstructured documents. Transforming these documents into digital format to perform various operations requires complex AI techniques on top of simple optical character recognition.

Today, a number of industries including medicine, finance, law, and real-estate rely on human-intensive processes to convert forms and other documents into digital formats. Be it patients history locked in medical record files of hospitals, mortgage applications, or tax forms, these transcriptions all requires effort and human intervention on top of simple OCR techniques — that used to detect text without keeping the composition of document intact — to convert them into valuable digital assets.

Continue reading Amazon Textract — Going beyond optical character recognition (OCR)

If there’s one thing that developers like more than a bug-free code, it’s a properly implemented dark theme! Be it on their IDEs or in the apps they use, dark theme is an integral part of the entire developer culture.

And while it not only looks good, it has several other benefits like:

With the introduction of Android 10, Google brought the much-awaited Dark Mode to the native Android framework and system apps.

Continue reading Adding Dark Mode Support to Your Android Apps

Depth sensing is a classic problem with a long history of prior work. It’s at the heart of many tasks, from 3D reconstruction to localization and tracking. Its applications span otherwise disparate research and product areas, including indoor mapping and architecture, autonomous cars, and human body and face tracking.

Continue reading ActiveStereoNet: The first deep learning solution for active stereo systems

The world is currently generating 2.5 quintillions of data each day, according to Forbes. The power of deriving meaning from these loads of data relies on the use of specialized machine learning algorithms that have been developed over time. We can teach machines to learn—or they can learn on their own.

Continue reading A guide to the top machine learning algorithms

A Convolutional Neural Network (CNN) is a multilayered neural network with a special architecture to detect complex features in data. CNNs have been used in image recognition, powering vision in robots, and for self-driving vehicles.

In this article, we’re going to build a CNN capable of classifying images. An image classifier CNN can be used in myriad ways, to classify cats and dogs, for example, or to detect if pictures of the brain contain a tumor. This post will be at an introductory-level, and no domain expertise is required. However, we assume that the reader has a basic understanding of Artificial Neural Networks (ANN).

Continue reading Convolutional Neural Networks: An Intro Tutorial

In the last couple of years, machine learning has opened up new horizons in a wide range of industries, with advanced use cases emerging: Facebook’s facial recognition, Netflix’s recommended movies, PrismaAI’s image style transfer, Siri’s voice recognition, Google Allo’s natural language processing, and the list goes on.

Continue reading 25 Open-Source Machine Learning Repos to Inspire Your Next Project

AWS Amplify is a fantastic framework that helps you develop your web or mobile applications quickly. Not only does it enhances your current tech stack, but it actually has many built-in features that you don’t have to worry about, especially when your app is in the development process.

Features include:

Beyond these features, Amplify can be integrated with most popular frontend frameworks like React, Vue, Angular, Ionic, React Native, or plain old vanilla JavaScript, if you’d like.

Continue reading User Authentication with Amplify in a React Native and Expo app

Mobile devices are limited in their resources compared to PCs. Some calculations can be executed on such limited-resources devices. For example, doing some calculations on an image or a couple of images. As the number of images increases, however, the device may run out of RAM.

If a deep neural network, for example, is to be trained on a large dataset with thousands of images, mobile devices are not a great option for training; rather, we need to us a machine with much more resources.

Continue reading Uploading images from Android to a Python-based Flask server

In this post, we’re going to dive deep into one of the easiest and most interpretable supervised learning algorithm — decision trees. Decision tree algorithms can be used for both classification and regression. We’ll be discussing how the algorithm works, it’s induction, parameters that define it’s structure, and it’s advantages and limitations.

Continue reading Understanding the Mathematics Behind Decision Trees

In this post, we’re going to dive deep into one of the most influential unsupervised learning algorithms—k-means clustering. K-means clustering is one of the simplest and most popular unsupervised machine learning algorithms, and we’ll be discussing how the algorithm works, distance and accuracy metrics, and a lot more.

Unsupervised learning is a type of self-organized learning that aids us in discovering patterns in our data related to various features. It is one of the three main categories of machine learning, along with supervised and reinforcement learning.

Continue reading Understanding the Mathematics behind K-Means Clustering