By Matthew Zeiler

Clarifai is excited to announce that Major General Barbara Fast, U.S. Army, Ret. has joined the company as senior advisor and member of its Public Sector Advisory Council.

Major General Barbara Fast, U.S. Army, Ret., is known for her military and commercial leadership while trailblazing new understandings of military intelligence. Her deep expertise in intelligence and cybersecurity, combined with over a decade leading federal defense businesses for commercial companies like Boeing and CGI, makes Her uniquely suited to support the U.S. …

By Jeff Toffoli

Spring is here and we are introducing important new features and improvements with Clarifai Release 7.4. Clarifai is democratizing AI and giving developers access to tools that can accelerate AI solutions at unprecedented speeds. We are adding new models, workflows and platform functionality that reduces the complexity of deploying AI solutions in the real world.

Learning from examples is better than creating “rules”

Let’s start with some good news. Just about every aspect of Computer Vision is getting better with Machine Learning vs. the old “rule-based” approach, where researchers were crafting equations by hand. Images are more easily classified, objects are quickly detected, poses are accurately estimated, and faces are recognized. Not only are the end results better, but the path to getting these results is predictable and repeatable -good news for anyone trying to incorporate this technology into their business.

Once a problem has been clearly defined, and meaningful metrics identified, a data scientist simply needs to provide examples for their model…

Recently, deep convolutional neural networks (DCNNs) have attracted much attention in remote sensing. Deep learning-based methods have been adopted for a variety of SAR images tasks, including object detection (automated target recognition), land cover classification, change detection, and data augmentation. In order to transfer CNN experience and expertise from optical to SAR, we need to understand the difference between optical imagery and SAR data [2].

SAR imagery provides information about what’s on the ground, but distortions and speckle make these images very different from optical images. Since SAR data is slightly less intuitive than optical data, it imposes a challenge…

ML in Modern Storytelling

With over a billion uploads and over 5 million writers contributing every month, evaluating and sorting stories has become impossible for Wattpad to do manually.

This becomes particularly evident when you consider how complex stories are to begin with. There are a diverse set of building blocks that make up a great story, including genre, grammar, tone, dialogue, sentence structure, setting, and characters to name a few.

This is where Wattpad relies on their proprietary “Story DNA” technology and AI. Inspired by Pandora’s music genome project, story DNA technology leverages machine learning to generate insights content from the world’s most…

By Jeff Toffoli

The Clarifai API offers you an encrypted gRPC channel, as well as an HTTPS+JSON channel for making requests. Why the different options? Learn about the many convenient benefits of using the Clarifai API built on gRPC.

With our newly released Model Mode, you can combine non-trainable model operators with trainable models into AI workflows for production.

Is normalization indispensable for training deep neural network?

When training a classifier or detector, batchnorm is always applied to help with solving the problem of vanishing/exploding variables, however batchnorm can lead to worse model performance if training with very small batch size. Limited by available gpu memory, if we need to train with small batch size, normally we’d adopt synchronized batchnorm so that statistics of batchnorm can be aggregated over multiple gpus to mitigate the small batch size problem, with a cost of longer training time. The synchronization between gpus is not as efficient, it lowers gpu utilization.

This paper tackles the problem of vanishing/exploding variables in…


Early search engines powered by the likes of Yahoo! and Alta Vista offered simple keyword matching technology that helped users find content on the web. These services would count up the number of times a given search term was present on a web page and then rank search results based on keyword frequency. This approach was later augmented by Google and others where search results relevance were improved by analysing the relationship between different websites.

With advancements in machine learning, new techniques have been developed that allow users to search for content without using keywords at all (Yang et…


The World’s AI Deep learning workspace for developers, data scientists, and business operators.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store