Apple Inc. will launch its HomePod smart speakers in Japan this summer at a retail price of ¥32,800, according to its website, although the exact date has yet to be announced. On sale since February 2018 in the United States and a number of other countries, the artificial intelligence-enabled speakers have attracted interest for their sound quality and integration with Siri, Apple's voice-controlled personal assistant that responds to simple questions and commands. The tube-shaped speakers are 17 centimeters high and 14 cm wide. Inc. and Google LLC have also each developed smart speakers.

In this special guest feature, Matt Field, President and Co-Founder of MakerSights, says there's no denying that AI and ML are starting to change the nature of the retail industry. One of its greatest benefits will be in how intelligent automation can be used by humans to augment their jobs. Drawing from his extensive background in both retail and technology as the former employee #1 at Birchbox, Matt is helping to transform the way brands develop and bring to market new products by combining voice of the customer software and predictive analytics. Through the MakerSights platform, leading apparel, accessories, footwear and home brands are able to de-risk their most important product decisions, making more of the products their customers want, and driving greater speed and profitability across their businesses. Artificial intelligence (AI) and machine learning (ML) are no longer just buzzwords of tech giants and consulting firms.

We present ARU, an Adaptive Recurrent Unit for streaming adaptation of deep globally trained time-series forecasting models. The ARU combines the advantages of learning complex data transformations across multiple time series from deep global models, with per-series localization offered by closed-form linear models. Unlike existing methods of adaptation that are either memory-intensive or non-responsive after training, ARUs require only fixed sized state and adapt to streaming data via an easy RNN-like update operation. The core principle driving ARU is simple --- maintain sufficient statistics of conditional Gaussian distributions and use them to compute local parameters in closed form. Our contribution is in embedding such local linear models in globally trained deep models while allowing end-to-end training on the one hand, and easy RNN-like updates on the other. Across several datasets we show that ARU is more effective than recently proposed local adaptation methods that tax the global network to compute local parameters.

Walmart has been surveilling its checkout registers using a computer vision technology called Missed Scan Detection to identify when items move past the scanner without having been scanned. As reported by Business Insider, the technology has been implemented in more than 1,000 stores across the US over the past two years, and it monitors both self-checkout kiosks and traditional registers managed by human cashiers. The system runs on cameras that watch as items move across the register. If an unusual activity occurs, such as an item moving into a bag without being scanned, a checkout attendant will be notified to take action. Missed Scan Detection was designed to help reduce theft and other losses, a problem that has cost US retailers up to $47 billion in 2017.

AI is helping retailers customize their offerings, create personalized experiences, and make shopping more convenient. Customer centricity -- putting your customer at the center of your strategy -- has long been considered the holy grail of retail marketing. In the digital age, customer-centricity revolves around data and smart technologies like artificial intelligence (AI). With the help of AI, companies collect as much data as they can about their customers' wants, needs, and preferences, and then apply it to customize their offerings, create personalized shopping experiences, and make the purchase process simpler and more convenient. An example of new tools available for understanding customer habits is the Personality Insights service made possible by IBM's AI platform, Watson.

In this study, we investigated multi-modal approaches using images, descriptions, and title to categorize e-commerce products on Amazon.com. Specifically, we examined late fusion models, where the modalities are fused at the decision level. Products were each assigned multiple labels, and the hierarchy in the labels were flattened and filtered. For our individual baseline models, we modified a CNN architecture to classify the description and title, and then modified Keras' ResNet-50 to classify the images, achieving F1 scores of 77.0%, 82.7%, and 61.0%, respectively. In comparison, our tri-modal late fusion model can classify products more accurately than single modal models can, improving the F1 score to 88.2%. Each modality complemented the shortcomings of the other modalities, demonstrating that increasing the number of modalities can be an effective method for improving the accuracy of multi-label classification problems.

An Echo Dot owner claims that Amazon's Alexa assistant has started calling him a's*******' whenever he asks the personal assistant to play him music. Micheal Slade, 29, was reportedly shocked when his Echo Dot speaker began to swear at him following his cancellation of his Amazon Prime subscription. The incident has reportedly left Amazon engineers puzzled -- with the tech firm offering Mr Slade gift cards and a year of free Prime membership in compensation. Software is available for the Echo Dot speaker that can make Alexa curse -- but it is unclear whether someone might have deliberately uploaded this to the device. An Echo Dot owner claims that Amazon's Alexa assistant has started calling him a's*******' whenever he asks the personal assistant to play him music Online business owner and Cwmbran, South Wales resident Michael Slade, 29, said that the trouble began the day after called Amazon to cancel his subscription to their Prime subscription service.

Jeremy Stanley is giving a talk, "How Instacart is Using AI to Create the Most Efficient Shoppers Ever," at the O'Reilly Artificial Intelligence Conference in San Francisco, September 17-20, 2017. Subscribe to the O'Reilly Data Show Podcast to explore the opportunities and techniques driving big data, data science, and AI. Find us on Stitcher, TuneIn, iTunes, SoundCloud, RSS. In this episode of the Data Show, I spoke with Jeremy Stanley, VP of data science at Instacart, a popular grocery delivery service that is expanding rapidly. As Stanley describes it, Instacart operates a four-sided marketplace comprised of retail stores, products within the stores, shoppers assigned to the stores, and customers who order from Instacart.

Retail chain Walmart says it uses AI face detection in cameras at over 1,000 store locations. It turns out, those "selfie cameras" at Walmart's checkout lines have been watching you back. The retail chain utilizes what it internally calls Missed Scan Detection, which helps detect when an item lands in a shopping bag without being scanned. The AI-powered visual scanners/cameras are placed at both self-checkout registers and ones run by store cashiers. SEE ALSO: Did Walmart Convince FedEx to Dump Amazon?

Without the tech infrastructure of Amazon or Walmart, it has been harder for small and medium-sized retailers to compete in an increasingly digital market. For starters, knowing how to manage inventory between physical stores and online can be challenging for retailers who lack access to the data insights of their larger competitors. But new developments in AI and machine learning are democratizing the use of data. "Five or 10 years ago, you really had no way out," said Kishore Rajgopal, founder and CEO at NextOrbit, an AI platform for inventory and price management. "Amazon was the king, Amazon and Walmart had all the data and the data scientists and you could never catch them -- but now you can."