Here's how I cut down my family's reliance on Alexa. Until recently, my smart home setup was in chaos. After years of testing, buying, and upgrading to the latest smart home gadgets in an attempt to make my life easier, it became a bloated mess that was actually making it more complicated. My Alexa, Google Home, and Apple Home apps were awash with dead devices, duplicates, and automations that simply didn't work. My Hue Bridge, trying desperately to tie it all together, was creaking at the seams.

Joining a startup accelerator program is an opportunity that many entrepreneurs vie for. With only 1,300 programs around the world, startup accelerators exclusively invest their time and resources into the most promising companies. A startup accelerator is an intensive, typically mentor-based, program that aims to expedite a young company's growth. Accelerators bring a plethora of knowledge and experience, which can be crucial for first-time founders and the success of their companies. Beyond the mentorship opportunities, startup accelerators can also connect founders to investors and help them with fundraising.

Annotating data for supervised learning is expensive and tedious, and we want to do as little of it as possible. To make the most of a given "annotation budget" we can turn to active learning (AL) which aims to identify the most informative samples in a dataset for annotation. Active learning algorithms are typically uncertainty-based or diversity-based. Both have seen success in image classification, but fall short when it comes to object detection. We hypothesise that this is because: (1) it is difficult to quantify uncertainty for object detection as it consists of both localisation and classification, where some classes are harder to localise, and others are harder to classify; (2) it is difficult to measure similarities for diversity-based AL when images contain different numbers of objects. We propose a two-stage active learning algorithm Plug and Play Active Learning (PPAL) that overcomes these difficulties. It consists of (1) Difficulty Calibrated Uncertainty Sampling, in which we used a category-wise difficulty coefficient that takes both classification and localisation into account to re-weight object uncertainties for uncertainty-based sampling; (2) Category Conditioned Matching Similarity to compute the similarities of multi-instance images as ensembles of their instance similarities. PPAL is highly generalisable because it makes no change to model architectures or detector training pipelines. We benchmark PPAL on the MS-COCO and Pascal VOC datasets using different detector architectures and show that our method outperforms the prior state-of-the-art. Code is available at https://github.com/ChenhongyiYang/PPAL

Proptech startup Built AI has closed a £2m seed round for its commercial real estate analytics platform. Built AI uses natural language processing – a subset of AI – to extract key data from property brochures uploaded by users. The startup says this information can help commercial real estate investors save time when assessing properties. The platform can then use this data to create a business plan with estimated rates of return for potential investments. The London-based startup was co-founded by Firoz Noordeen, Natan Lempert and Jeffrey Ng.

Edge computing is booming, with estimates ranging up to $61 billion in value in 2028. While definitions vary, edge computing is about taking compute power out of the data center and bringing it as close as possible to the device where analytics can run. The devices can be standalone IoT sensors, drones, or autonomous vehicles. Increasingly, data generated at the edge are used to feed applications powered by machine learning models," stated George Anadiotis, analyst, engineer and founder of Linked Data Orchestration of Berlin, Germany, working on the intersection of technology, media and data, writing in a recent account in ZDnet. However, "There's just one problem: machine learning models were never designed to be deployed at the edge.

Text autoencoders are commonly used for conditional generation tasks such as style transfer. We propose methods which are plug and play, where any pretrained autoencoder can be used, and only require learning a mapping within the autoencoder's embedding space, training embedding-to-embedding (Emb2Emb). This reduces the need for labeled training data for the task and makes the training procedure more efficient. Crucial to the success of this method is a loss term for keeping the mapped embedding on the manifold of the autoencoder and a mapping which is trained to navigate the manifold by learning offset vectors. Evaluations on style transfer tasks both with and without sequence-to-sequence supervision show that our method performs better than or comparable to strong baselines while being up to four times faster.

Fintech Europe, Plug and Play's fintech-focused innovation platform based out of Frankfurt, Germany, announced today the eight startups selected for its sixth batch. The platform has grown its partner base to 13 Financial Institutions since its inception in May 2018. Together with Deutsche Bank, TechQuartier, BNP Paribas, Nets Group, Nexi, UniCredit, Aareal Bank, Abanca, Danske Bank, DZ Bank, Elo, UBI Banca, and Raiffeisen Bank International, the program seeks to support innovation in the world of Financial Services. After screening applications from all over the world and intensive weeks of reviewing preselected startups with the partners, the final group of eight companies have been accepted into Fintech Europe. The program aims at facilitating pilots, POCs, and business development opportunities for the participating startups and financial institutions.

The tools, announced at Amazon's annual cloud event in Las Vegas, aim to help the company's cloud customers incorporate functionality such as natural language processing, but without long waits often associated with AI-related projects. Amazon says it is making this possible by integrating machine learning into the two new stand-alone services. "There's no machine-learning expertise required for either of these services. They're just plug and play. You don't have to get into all the weeds and get the training data and label the data and all those sorts of things," said Matt Wood, vice president for artificial intelligence services at Amazon Web Services.

Willis Towers Watson, a leading global advisory, broking and solutions company, and Relativity6, Inc., a machine learning and artificial intelligence (AI) insurance-technology company, today announced that Willis Towers Watson has selected the Relativity6 platform to predict and optimise customer retention and win-back. Brent Lehmann, General Manager Affinity & Commercial Australasia said the partnership with such an innovative technology company will help to ensure Willis Towers Watson remains competitive in the marketplace. "Relativity6's product offerings are a good fit to accomplish our strategic objectives across the organisation, so we are very excited to partner with them to take full advantage of the data that we have accumulated within our core systems in Australia." Alan Ringvald, Chief Executive Officer at Relativity 6, commented: "We are honoured to partner with such a distinguished organization. We believe that our solution will enable Willis Towers Watson to better serve their customers and ultimately drive significant top line revenue growth. We've engaged with top-tier insurers in the U.S. and Latin America, and this is a fantastic opportunity to expand our footprint with a truly global insurance broking brand."

The banking industry is changing. In the words of sci-fi author William Gibson, "the future is already here – it's just not evenly distributed." In China, a payments revolution is underway that constitutes a true "Great Leap Forward." While the U.S. is still mapping the path to real-time payments, and scrutinizing the disappointing penetration of mobile payments, China is well on the way to being a cardless, as well as a cashless, society. Nearly 50% of WeChat's billion users have now paid at a retail store using a QR code, while Alipay (with another 500 million users) is seeking to bypass the phone entirely with its "pay with a smile" facial recognition system.