Microsoft Unveils Lobe; Will this Make AI Mainstream?
New Lobe software enables anyone to create and train machine learning models.
Posted Oct 26, 2020
The complexity of artificial intelligence (AI) machine learning requires specialized knowledge and experience. Microsoft hopes to change that. Today Microsoft provided a general preview of Lobe, a software application available at no charge that enables anyone to build machine learning models—no technical skills required.
Recent trends such as decentralized cloud computing, adaptation of GPU for general computing, increasing availability of big data sets, and advances in deep learning, a subset of AI machine learning, has spurred a modern-day AI gold rush.
Global investment in AI in just half a decade has soared across sectors and geographies. Privately held AI companies garnered roughly USD 40 billion in investment funding spread over 3,100 transactions worldwide in just last year alone, out of which American startups account for an estimated 64 percent of the global share or USD 25.2 billion and 1,412 transactions, according to Georgetown University Center for Security and Emerging Technology (CSET) September 2020 report. These are conservative figures as the actual total transaction value globally for 2019 could be as high as USD 74 billion when factoring in CSET estimates for transactions without publicly disclosed values.
Lobe was founded in August 2016 with headquarters in San Francisco, California according to Bloomberg. The company founders include Mike Matas, Adam Menges, and Markus Beissinger. Lobe had venture-capital funding from Chris Sacca’s Lowercase Capital according to PitchBook. In September 2018, Microsoft acquired AI startup Lobe in efforts to enable anyone to perform artificial intelligence development.
Lobe is a software application that resides on Windows or Mac desktop computers that enables anyone to create machine learning models for image classification. The steps are simple—create a dataset using a web camera or existing pictures, label the categories, train the model, review results, then run the model.
Once the model is created, it can be exported to run on multiple platforms. For example, Lobe models can be exported as TensorFlow 1.15 SavedModel, a standard format used in Python applications that run TensorFlow 1.x, or hosted on AWS, Google Cloud and Azure. Lobe also supports Apple iOS via Core ML to develop iOS, iPad and Mac apps. Exports to TensorFlow Lite support mobile and IoT applications for Android or Raspberry Pi. Lobe supports local API, spreadsheets and local images. Lobe provides APIs for Python and .NET for exported models.
The uses of Lobe for image classification are vast. Examples include training the app to distinguish toxic from non-toxic plants, react to facial expressions with emojis, and validate proper mask-wearing for safety. Currently Lobe supports projects for image classification, with plans to release templates for object detection and data classification in the future.
With the release of Lobe to the general public, Microsoft has taken the first step in bringing machine learning to the masses—a move that may accelerate AI crossing the chasm from B2B to mainstream customers. Now anyone has access to machine learning capabilities and can create models without requiring any technical knowledge—a new milestone towards the democratization of artificial intelligence.
Copyright © 2020 Cami Rosso All rights reserved.