For years now, Snapchat has been at the forefront of mobile machine learning — their popular Lenses, which often combine on-device ML models with augmented reality, have become shining examples of the power and flexibility of on-device machine learning.
Given our respect and admiration for Snap’s work in this area, our team was thrilled to hear about the recent release of SnapML, Snap’s new ML framework inside their development platform Lens Studio (released with 3.0).
Last year, at I/O 2018, Google announced a brand new SDK available for developers: ML Kit. It’s no surprise that Google’s advances in machine learning are miles ahead of what any other company is aiming for. Through this SDK, Google was hoping to help mobile developers bring machine learning to their apps with simple, concise code. As part of the Firebase ecosystem, ML Kit allows developers to implement ML functionality with just a few lines of code; everything from vision to natural language to custom models.
Continue reading Exploring the new ML Kit features on iOS using Swift
One thing I’ve observed in many data science tutorials when it comes to modeling, is that once a certain performance threshold is achieved on test data, rarely is the model deployed/pushed to production—and it’s a common case in the industry more broadly.
This tutorial aims to take modeling a step further by building a REST API and deploying the model into production. In addition to the REST API, we’re building a simple web application that predicts whether a piece of text belongs to any of these classes: atheism, computer graphics, medical science, Christianity, or politics.
Continue reading Deploying and Hosting a Machine Learning Model Using Flask, Heroku and Gunicorn
You must’ve heard about “Vin Rouge” by @Nikhilodeon12 by now. The newly-verified Snap Star Nikhil created this lens, sparking the viral “Silhouette” challenge on TikTok. That’s right, TikTok.
The platform filled with its own filters, special effects, and AR tools. How did a Snapchat Lens end up over there?
Continue reading “How Snapchat Lenses affect TikTok trends — and why Lens Creators are so important”
In my article on lessons from my first deep learning hackathon, I introduced heatmaps. I mentioned that they highlight the regions in an image that the CNN focuses on while trying to make a prediction, and that it would be interesting to learn how they’re generated. Well, I finally know how. Let’s first go back and see what they look like.
In this post, we’re going to investigate the field of image compression and it’s applications in real world. We’ll explore various machine and deep learning techniques for image compression and inspect their pros and cons, and their practical feasibility in real-world scenarios.
So let’s get started!
Image compression refer to reducing the dimensions, pixels, or color components of an image so as to reduce the cost of storing or performing operations on them. Some image compression techniques also identify the most significant components of an image and discard the rest, resulting in data compression as well.
Continue reading Image Compression Using Different Machine Learning Techniques
Google Colab is a cloud-based service that allows the execution of Python code and includes the ability to use and install new libraries. Put simply, you can leverage Google Colab to create a Jupyter Notebook and execute Python code using either a CPU or a free GPU.
Continue reading Getting Started with Google Colab Notebooks
Recently, there has been a huge rise in the implementation of artificial intelligence solutions, with new deep learning architectures being built and deployed across various industries. This rise could be attributed to two important factors:
Deep learning works primarily because of the vast amount of input data on which the deep neural net is trained. Hence, having a good labeled training dataset marks the first step in developing a highly accurate AI solution.
Continue reading Image Manipulation for Machine Learning in R
Neural network implementation became a lot easier since the advent of transfer learning in accessible libraries. So much so that deep learning code that previously required hours to write can be written today in just 2 lines — No kidding !
Let me demonstrate what transfer learning is through the most basic example — our daily lives.
Remember when you last purchased a new phone — and didn’t even have to spend a day learning how it works ? As humans, we learn based on previous experiences. When we transitioned to our second smartphone, we already had mental and behavioral models to bring us up-to-date with the new device. This is what is known as transfer learning in the domains of data science and machine learning . So, what is this transfer learning we practice so much and know so little about?
Continue reading Image Segmentation with Transfer Learning [PyTorch]
There are over 5 billion mobile device users all over the world. Such users generate massive amounts of data—via cameras, microphones, and other sensors like accelerometers—which can, in turn, be used for building intelligent applications. Such data is then collected in data centers for training machine/deep learning models in order to build intelligent applications.
However, due to data privacy concerns and bandwidth limitations, common centralized learning techniques aren’t appropriate—users are much less likely to share data, and thus the data will be only available on the devices.
