HullBreach Studios has published a new app on Apple stores today under the name “ML Image Identifier” (It shows as “ML Image ID” under the icon)! “ML” stands for Machine Learning and is compiled making use of ML Models, which is a reference to the structure of the “brain”. It’s a file format.. that defines the neurons and the weights of their connections, based on what they have “learned”.. from test data.
Xcode 10 makes it really easy to add them to any project in development. You just drag-and-drop them, and it’s immediately available to the code. Something that used to need a cluster of supercomputers is now possible on a common mobile device! The new app “ML Image Identifier” is basically considered as an educational app in that it allows your iOS device to identify images (in real-time), as you move the camera around your 3D environment. It scans for 3 categories images included in the launch version of ML Image Identifier Released today: “objects”, “cars”, and “food”, each with various levels of accuracy (high, medium, and low, respectively). It automatically throttles the image processing to work on any device running iOS 11. The top-5 predicted matches are listed, based on the neural networks’ confidence levels as percentages. This ML app needs iOS 11.4 and newer. That’s iPhone 5S and newer, iPad Air and newer, iPod Touch 6, any of the iPad Pro models, or iPad Mini 2 and newer, basically, any of the 64-bit devices from the past 5 years.
Once merely a subject of science-fiction, machine learning has permeated our lives in recent decades. We see it in numerous uses, such as handwriting recognition, facial recognition, image tagging, AI in games, targeted advertisements, predictive typing, and many automated tasks. Social networks are free because the data you provide can be valuable for numerous purposes. In short: Knowledge is power. With the release of iOS 11, Apple brought machine learning to the masses with CoreML, making it possible to run neural networks and other ML-related tools via hardware acceleration on any iOS device. This app here is a demonstration of some real possibilities and some deficiencies of machine learning. Modeling a neural network is only one part of the task. For a ML model to work, it must be fed massive amounts of test data (similarly to how it takes a living creature numerous stimuli to learn). Good test data can yield good results; poor test data can yield poor results. Sometimes, biases of those creating the tests can come into play, since they may unknowingly weight certain tests values over others.
ML Image Identifier App makes use of 3 ML models
#1= “MobileNet” This scans general objects. It works fairly well with household items. It cannot identify people. This ML model is an example of fairly high-quality results in image recognition and is much more compact than similar ML models that can be as large as 500MB.
#2= “CarRecognition” This scans for makes and models of vehicles. It is very hit-or-miss and seems to heavily match automobiles from specific regions. Most matches are the right body type but wrong make. This ML model is an example of mixed results in image recognition.
#3= “Food101” This scans for prepared foods. It rarely works with general food items and seems to focus on foods that most people will not have in their houses, such as caviar and lobster. It also returns many false-positives for desserts. This ML model is an example of poor results in image recognition.
Post Sponsored By:
Pixel Pixie Productions