Samsung Gaming Hub goes live today with Twitch, Xbox Game Pass and more

The Samsung Gaming Hub is live now on 2022 Samsung smart TVs and smart monitors, and it’s adding two services from Amazon to its game-streaming lineup: Twitch and Luna. Twitch is available today, while Luna is coming soon. Gamers will also be able to access Xbox Game Pass now, as well as apps for NVIDIA GeForce NOW, Google Stadia and Utomik in the same designated area on their TVs. The company plans to release details about the gaming hub’s rollout to earlier Samsung smart TV models at a later date, a Samsung spokesperson confirmed to Engadget. 

For those who are unfamiliar with the Samsung Gaming Hub, it essentially offers players a way to access major cloud gaming services on their smart TV using only their Bluetooth controller, . Apps for both Spotify and YouTube are also included in the gaming hub.

Samsung says it plans on delivering even more gaming-focused content in the future, including new partnerships. “With expanding partnerships across leading game streaming services and expert curated recommendations, players will be able to easily browse and discover games from the widest selection available, regardless of platform,” said Won-Jin Lee, president of Samsung’s Service Business Team.

Amazon’s Luna cloud gaming service has only been to the general public since March, and is already available on Fire TVs. Its partnership with Samsung could give the nascent gaming service an easy way to reach people who have never used it in their homes. Twitch (which is owned by Amazon) once had an app for Samsung smart TVs, but it was in 2019. The platform’s return to the newest Samsung smart TVs will be happy news for streamers and their fans.

It seems natural for Samsung to further embrace the gaming community, given that smart TVs have become close to a necessity in gaming. Last year Microsoft announced that it would begin working with global TV manufacturers to directly Xbox into smart TVs via an Xbox with Game Pass app. The idea of an “all-in-one” destination for all your cloud-based and console games is certainly convenient to some, and may help gamers avoid the time and hassle of switching between modes.

All products recommended by Engadget are selected by our editorial team, independent of our parent company. Some of our stories include affiliate links. If you buy something through one of these links, we may earn an affiliate commission.

Repost: Original Source and Author Link


Snapchat Plus is now live, and it costs $4 per month

Snapchat Plus has officially launched today, and like Twitter’s own paid subscription, Twitter Blue, it appears to be designed for power users of the app, those who would be most interested in all the experimental and exclusive features that Snapchat has to offer.

On Wednesday, Snapchat announced the release of Snapchat Plus via a blog post. Few details about the new paid subscription were given in the announcement, but here’s what we do know:

  • It costs $4 per month and there aren’t different payment tiers as previously reported.
  • Snapchat Plus is expected to offer “a collection of exclusive, experimental, and pre-release features available in Snapchat.” So like Twitter Blue, it offers subscribers access to features and app customizations for a monthly fee.
  • It is expected to launch today in the following countries:  the U.S., Canada, the U.K., France, Germany, Australia, New Zealand, Saudi Arabia, and the United Arab Emirates.
  • You should be able to sign up for Snapchat Plus via the app by selecting Snapchat+ on your profile. However, when we tested this on an Android device, that option didn’t appear to be available yet. It is possible that Snapchat Plus is still rolling out to devices in the countries listed above.

In terms of features, The Verge reports that Snapchat Plus will offer little more than “cosmetic” changes: You’ll be able to alter the style of the app icon, pin and label a friend as a “BFF,” and “see who rewatched a story.”

And unfortunately, like Twitter Blue, The Verge notes that even if you pay for a subscription to Snapchat Plus, that doesn’t buy you an ad-free experience. You’ll still be stuck with ads even after shelling out $4 per month.

Editors’ Choice

Repost: Original Source and Author Link


Tesla AI Day event: start time and how to watch the live stream

Today is Tesla’s AI Day, a sequel of sorts to the company’s Autonomy Day event held in 2019. The event, which will be held at Tesla’s headquarters in Palo Alto, CA, will be livestreamed for the public starting at 5PM PT / 8PM ET (though the event may not actually begin until closer to 5:30PM PT).

We don’t have a lot of details about what will be announced, but based on the invitation, we’ll get a keynote address by Tesla CEO Elon Musk, hardware and software demos from Tesla engineers, test rides in the Model S Plaid, and “more.” Musk has also tweeted that the “sole goal” of the event is to lure experts in the field of robotics and artificial intelligence to come work at Tesla.

Tesla usually holds around two public events a year. This year, we got the Tesla Model S Plaid launch event in June, and now AI Day. Over the past few years, Tesla has been holding events, not to unveil new products, but to highlight certain technologies that the company views as crucial to its future development. Last year, Tesla held its first Battery Day event, at which it discussed plans to drive down the cost of battery development with the goal of producing a $25,000 electric car.

AI Day is expected to pick up on the themes first introduced during Autonomy Day, which include the manufacturing of Tesla’s own silicon computer chips to power it’s Full Self-Driving advanced driver assistance feature.

This AI Day comes at an awkward time for the company. Earlier this week, the National Highway Traffic Safety Administration announced that it was investigating Tesla’s Autopilot for the nearly dozen incidents in which its cars crashed into emergency vehicles. Two Democratic senators also called on the Federal Trade Commission to investigate Tesla’s marketing practices for potentially misleading information.

Repost: Original Source and Author Link


How to watch The Game Awards 2021 streaming live

Today is a big day for the games industry, as The Game Awards is returning tonight. Not only will the show dole out awards to a variety of games across many categories, but we’ll also be getting some new game reveals as well. In short, there’s going to be a little something for everyone, even those who aren’t interested in awards shows in general.

How and when to watch The 2021 Game Awards

Luckily for those interested in watching, The Game Awards traditionally livestreams on a large number of platforms. The easiest way for most of us to watch will probably be on YouTube (via the livestream we’ve embedded below) or Twitch. On YouTube, the show will be streaming in 4K, while Twitch will offer a special Game Awards extension that will allow viewers to interact with the stream. On both platforms, co-streaming is allowed, so you’ll be able to watch other streamers co-stream the event too.

If YouTube and Twitch aren’t quite your style, there are more platforms beyond those that will be hosting the show. The Game Awards website lists platforms like Facebook, Twitter, IGN, GameSpot, and TikTok as global streaming partners, so there are many more places to watch beyond just Twitch and YouTube. The Game Awards also has a number of streaming partners worldwide, so if you live outside of the US, it might be worth checking that page we just linked to see if there are any platforms hosting the stream in your region.

While the main show doesn’t kick off until 8 PM EST tonight, there will be a pre-show beginning at 7:30 PM EST. While the biggest award of the night is undoubtedly Game of the Year – the nominees for which include Deathloop, It Takes Two, Metroid Dread, Psychonauts 2, Ratchet & Clank: Rift Apart, and Resident Evil Village – there are 30 awards in all that span gaming and gaming-adjacent categories.

What we expect at The Game Awards this year

Aside from the obvious awards ceremony, The Game Awards also serves as a place for developers to reveal their upcoming games. The description on that YouTube livestream we’ve embedded above gives away a couple of those reveals, noting that Jim Carrey and Ben Schwartz will be on hand to reveal Sonic the Hedgehog 2 (the movie, not the game from 1992), while Keanu Reeves and Carrie-Anne Moss will debut The Matrix Awakens.

The Game Awards host and producer Geoff Keighley has been making the rounds hyping up his show, saying in a recent interview with ForTheWin that there will be several game announcements on the level of the Elden Ring reveal we saw over the summer.

In a recent AMA on Reddit, Keighley also said that the show should last around three hours, so plan on settling in if you want to watch the show from start to finish. He also confirmed that there will be some surprise guests at The Game Awards, so we’re in for at least a couple of surprises. We don’t have much longer to wait before The Game Awards is underway, so tune in to one of the livestreams we’ve linked above at 7:30 PM EST, 4:30 PM Pacific if you want to watch the show play out live tonight (December 9, 2021).

Repost: Original Source and Author Link


Meta launches PyTorch Live to build AI-powered mobile experiences

Join gaming leaders, alongside GamesBeat and Facebook Gaming, for their 2nd Annual GamesBeat & Facebook Gaming Summit | GamesBeat: Into the Metaverse 2 this upcoming January 25-27, 2022. Learn more about the event. 

During its PyTorch Developer Day conference, Meta (formerly Facebook) announced PyTorch Live, a set of tools designed to make AI-powered experiences for mobile devices easier. PyTorch Live offers a single programming language — JavaScript — to build apps for Android and iOS, as well as a process for preparing custom machine learning models to be used by the broader PyTorch community.

“PyTorch’s mission is to accelerate the path from research prototyping to production deployment. With the growing mobile machine learning ecosystem, this has never been more important than before,” a spokesperson told VentureBeat via email. “With the aim of helping reduce the friction for mobile developers to create novel machine learning-based solutions, we introduce PyTorch Live: a tool to build, test, and (in the future) share on-device AI demos built on PyTorch.”

PyTorch Live

PyTorch, which Meta publicly released in January 2017, is an open source machine learning library based on Torch, a scientific computing framework and script language that is in turn based on the Lua programming language. While TensorFlow has been around slightly longer (since November 2015), PyTorch continues to see a rapid uptake in the data science and developer community. It claimed one of the top spots for fast-growing open source projects last year, according to GitHub’s 2018 Octoverse report, and Meta recently revealed that in 2019 the number of contributors on the platform grew more than 50% year-over-year to nearly 1,200.

PyTorch Live builds on PyTorch Mobile, a runtime that allows developers to go from training a model to deploying it while staying within the PyTorch ecosystem, and the React Native library for creating visual user interfaces. PyTorch Mobile powers the on-device inference for PyTorch Live.

PyTorch Mobile launched in October 2019, following the earlier release of Caffe2go, a mobile CPU- and GPU-optimized version of Meta’s Caffe2 machine learning framework. PyTorch Mobile can launch with its own runtime and was created with the assumption that anything a developer wants to do on a mobile or edge device, the developer might also want to do on a server.

“For example, if you want to showcase a mobile app model that runs on Android and iOS, it would have taken days to configure the project and build the user interface. With PyTorch Live, it cuts the cost in half, and you don’t need to have Android and iOS developer experience,” Meta AI software engineer Roman Radle said in a prerecorded video shared with VentureBeat ahead of today’s announcement.

Built-in tools

PyTorch Live ships with a command-line interface (CLI) and a data processing API. The CLI enables developers to set up a mobile development environment and bootstrap mobile app projects. As for the data processing API, it prepares and integrates custom models to be used with the PyTorch Live API, which can then be built into mobile AI-powered apps for Android and iOS.

In the future, Meta plans to enable the community to discover and share PyTorch models and demos through PyTorch Live, as well as provide a more customizable data processing API and support machine learning domains that work with audio and video data.

PyTorch Live

“This is our initial approach of making it easier for [developers] to build mobile apps and showcase machine learning models to the community,” Radle continued. “It’s also an opportunity to take this a step further by building a thriving community [of] researchers and mobile developers [who] share and utilize pilots mobile models and engage in conversations with each other.”


VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Repost: Original Source and Author Link


Epic Games Store Black Friday sale live: The best deals on games to seek

Black Friday is just a couple of days away, so of course, we’re seeing many digital storefronts kick off their Black Friday sales this week. One such store is the Epic Games Store, which launched its Black Friday sale just a short time ago. Epic is getting the jump on Steam here, launching its Black Friday sale just a few hours before Valve is expected to kick off the Steam Autumn Sale later today.

Big games on sale, but not for long

These Black Friday sales can be seen as something of a precursor to the more extensive winter sales we’re bound to see toward the end of December. For instance, Epic’s Black Friday sale only lasts from today until November 30th, meaning it’ll be live for less than a week. By comparison, the winter sale will probably last a couple of weeks at a minimum to close out 2021 and usher in 2022.

We can probably expect the same from Steam, so these Black Friday sales are really just a primer for sales to come. With that said, if you see a good deal on a game you want during this Black Friday sale, it’s probably a good idea to jump on it now instead of waiting for the larger winter sale to roll around next month. After all, there’s no guarantee that the game will get a similar discount during the next sale, likely though it may seem.

Recommended deals to check out

There are a surprising number of recently released, big-name titles getting discounts in Epic’s Black Friday sale. Some of the highlights include Far Cry 6 for $49.79 (17% off), Back 4 Blood for $41.99 (30% off), and the Crysis Remastered Trilogy for $39.99 (20% off). Epic says this is the first time any of those games have been discounted, and given that they can still be considered new to some extent, we believe it.

Other discounts that stick out to us include Sega’s 4x strategy game Humankind for $39.99 (20% off) and Darkest Dungeon II – a game that only recently launched in early access – with a 10% discount that knocks its price down to $26.99. In addition, Borderlands 3 and Grand Theft Auto V Premium Edition are both down to $14.99, while Horizon Zero Dawn: Complete Edition is half off at $24.99.

The list containing all of the deals stretches on for 12 pages, so there are a ton of deals to sift through. The Epic Games Black Friday sale runs until November 30th, so check out the deals for yourself and see if anything strikes your fancy.

Repost: Original Source and Author Link


Why fast, effective data labeling has become a competitive advantage (VB Live)

Presented by Labelbox

Iterating on training data is key to building performant models, but perfecting and tightening the loop still remains a challenge for even the most advanced teams. For practical insights on how to get models to production-level performance quickly with high-quality training data, don’t miss this VB Live event.

Register here for free.

The greatest challenge faced by machine learning engineers today is the number of time-consuming steps between gathering data and having a high-performing model. These steps can be incredibly laborious, and many ML teams in enterprises lack the infrastructure or tools to do it quickly enough.

“One of the biggest learnings we’ve had over the last few decades as a community is that the cornerstone for success in technology and engineering is faster iterations,” says Manu Sharma, CEO & cofounder of Labelbox. “The reason leading AI companies are successful is they’re iterating fast. They learn from each cycle and they improve rapidly.”

Most teams, however, don’t have the streamlined workflows or the right tools to move quickly enough to get their models into production on the timeline they want.

The biggest challenges for ML teams

Almost every enterprise-sized company now has goals to integrate AI into some aspects of their business, from finance to marketing to customer service — enabling more automation, smoother processes, and new products and services that were previously impossible. Getting to high-performing AI, however, is often hindered by several challenges.

For a company making AI-based products that will work across many different geographical regions or environments, their models have to be extremely accurate and robust. To build them, teams need to train and test models repeatedly, which in turn requires a vast amount of training data across a wide variety of scenarios, as each model needs to be tested successfully against each scenario.

Even teams with AI models in production need to constantly retrain and refresh them with new data. Because these models are so hungry for data, the number-one bottleneck for iterating with these models is data labeling. The most common way to handle it is outsourcing — which is a valid choice — but there are ways to improve the way it’s done now. Data labeling can be optimized using a training data platform: software that enables transparent communication and collaboration between machine learning engineers, domain experts, and outsourced teams, so that they can uncover problems and fix them right away in an iterative process.

The other big challenge for ML teams is the process of identifying and adjusting labels and training data for edge cases. Depending on the use case, data sources, and other variables, the number of edge cases can be large. To identify them quickly during the training process, it’s important for training datasets to be diverse and represent as many real-life situations as possible.

Teams can use automation to help discover these edge cases, figure out which ones are important, which ones are not, and then  work precisely to solve those problems. “Problems are solved by labeling more data that resembles those edge cases, because the model needs to see more examples,” says Sharma.

Take for instance self-driving AI models. A human driver can instantly make decisions about most unexpected situations while they’re driving, from a child running across the street to wet pavement from rainfall. An AI tasked with the same hurdles needs to be trained on data that represents every possible scenario that a driver can face.

Or consider home rental organizations that need to verify that all listings are legitimate. Having a person verify all the photos that users upload can be expensive and unwieldy, so some companies have developed AI models to automatically judge whether a photo’s description matches the picture and flag misinformation. But again, the number of edge cases can dramatically affect how the algorithm performs.

Tackling the challenge

If an AI model can make decisions on the company’s behalf through products and services, that model is essentially their competitive edge — and its performance entirely depends on the quality of the labeled data that was used to train it. Business leaders should think of training data as a competitive advantage and prioritize its quality and cultivation.

There is no silver bullet, however: the primary way for ML teams to break through bottlenecks and speed up innovation is to invest in infrastructure — including the tools and the workflows that enable ML teams to turn datasets into labeled data and make use of it. These tools should make it easy for teams to bring together every part of their labeling pipeline into a seamless process, including sending datasets to labelers, training labelers on the ontology and use case, quality management and feedback processes, model performance metrics that identify edge cases, and more.

“Choosing the right technology inherently brings the stakeholders together and streamlines their workflows and processes,” Sharma says. “By virtue of that, business leaders should be asking their teams to choose the right technologies to foster collaboration and transparency.”

To learn more about how to speed up the iteration cycle, label data quickly and effectively improve your competitive advantage, and how to choose the right tools and technology, join this VB Live event.

Register here for free.

You’ll learn how to:

  • Visualize model errors and better understand where performance is weak so you can more effectively guide training data efforts
  • Identify trends in model performance and quickly find edge cases in your data
  • Reduce costs by prioritizing data labeling efforts that will most dramatically improve model performance
  • Improve collaboration between domain experts, data scientists, and labelers


  • Matthew McAuley, Senior Data Scientist, Allstate
  • Manu Sharma, CEO & Cofounder, Labelbox
  • Kyle Wiggers (moderator), AI Staff Writer, VentureBeat

Repost: Original Source and Author Link


How proptech is removing friction from the home-buying process (VB Live)

Presented by Envestnet | Yodlee

A new wave of fintech is penetrating a wide range of financial processes, thanks to the accelerating adoption of digital solutions during the pandemic. Join this VB Live event to learn more about the tremendous opportunities available for entrepreneurs.

Registration here for free.

The pandemic kicked off a tremendous amount of uncertainty in the real estate industry, initially considered a non-essential business, engendering a lot of soul-searching for companies. While real estate technology platforms have been coming into their own for some time, the pandemic marked an urgent turning point. Agents had to figure out how to turn a business that’s typically very hands-on and personal — from home tours to document signing — into a much more resilient, efficient virtual business.

Luckily, fintech is also at a turning point, where the underlying, low-level APIs and services and product suites are available off the shelf to power innovative new platforms, or property technology. Also known as proptech, the term generally refers to the application of information technology and platform economics to real estate markets including how people research, rent, buy, sell, and manage a property.

The groundwork for these solutions has been laid over the past 10 to 15 years by the larger fintech ecosystem, bringing the APIs and tools developers need to build the next generation of consumer-focused tech to life. Global funding in proptech rose from $856 million to $1.76 billion from January to March this year.

“Real estate transactions are fraught with a ton of contingency and uncertainty. We’re looking to untangle that by owning the full end-to-end transaction,” says Joe Mocerino, vice president of engineering, at HomeLight. “The fact that you don’t have to build some of this tech, and you can just buy it off the shelf and plug it into your platform, it’s incredible.”

And the more that HomeLight is able to automate with technology in order to make stronger, quicker decisions, the more they have to offer customers, Mocerino says. They’re able to identify the agents in any given market and match them with consumers, enable mortgages, title and escrow, closing services, and other parts of the real estate transaction, and come up with innovative new ways to engage those products, he says.

There are downstream benefits, besides the ability to improve their services for consumers: leveraging respected technology also gives more confidence to the secondary banks that ultimately help fund these loans.

“When they see the use of this technology done in a very systematic way, there’s more trust when it comes to lending money to us to fund these loans,” he explains.

The data question

Algorithms and automation are essential, but require human control, or checks and balances, because data quality control should be a company’s fundamental consideration. Mocerino points to the company’s mortgage offering, and their loan origination system, which uses operator software to underwrite a borrower.

“Real estate data is historically incredibly messy, coming from MLS systems that are not necessarily the most tech-forward,” he explains. “We tap into these technologies to bring the data in, but we still have a human review it and verify it.”

The company buys overlapping data sets from multiple vendors whenever possible, and then tests and verifies, on both the financial and real estate sides. Their data team is careful not to take too much risk with the data, analyzing it before presenting it to a user or plugging it into a transaction.

The future of proptech

HomeLight’s vision for the future is a contingency-free transaction, making it as easy as connecting a buyer and the seller who are both qualified and making the sale, without multi-step contingencies, as well as growing its a trade-in and cash offer products, HomeLight Trade-In and HomeLight Cash Offer, which helps consumers compete in an increasingly buy market.

“I’m personally very excited about where underlying fintech products, like APIs and services, are going, and what that means for future products that aren’t built yet,” he says. “Entrepreneurs are seeing these tools that they haven’t had in the past, and I think there are all types of new categories of products that will let you be innovative on the technology side.

To learn more about how new segments of fintech are exploding onto the market, the opportunities for entrepreneurs and developers alike, and more, don’t miss this VB Live event!

Registration is free here.

Attendees will learn:

  • How the fintech ecosystem is evolving — and how fintech in a box is changing the landscape for developers
  • The most important APIs to avoid reinventing the wheel
  • How to ensure your data is accurate, reliable and diversified — and why that’s important
  • What the fastest growing new fintech segments are
  • And more!


  • Ran Harpaz, Chief Technology Officer, Hippo Insurance
  • Joe Mocerino, Vice President of Engineering, HomeLight
  • Seb Taveau, Head of Developer Experience | Center of Excellence | Envestnet
  • Seth Colaner, Moderator, VentureBeat

Repost: Original Source and Author Link


Designing the optimal iteration loop for AI data (VB Live)

Presented by Labelbox

Looking for practical insights on improving your training data pipeline and getting machine learning models to production-level performance fast? Join industry leaders for an in-depth discussion on how to best structure your training data pipeline and create the optimal iteration loop for production AI in this VB Live event.

 Register here for free.

Companies with the best training data produce the best performing models. AI industry leaders like Andrew Ng have recently emerged as major proponents of data-centric machine learning for enterprises, which requires creating and maintaining high-quality training data. Unfortunately, the tremendous effort it takes to gather, label, and prep that training data often overwhelms teams (when the task is not outsourced) and can compromise both the quality and quantity of training data.

Just as importantly, model performance can only improve at the speed at which your training data improves, so fast iteration cycles for training data is crucial. Iteration helps ML teams find new edge cases and improve performance. Additionally, iteration helps to refine and course correct data throughout the AI development lifecycle to maintain its reflection of real-world conditions. Shrinking the length of that iteration cycle lets you hone your data and conduct a greater number of experiments, accelerating the path to production AI systems.

It’s clear that iterating on training data is vital to building performant models quickly — so how can ML teams create the optimal workflow for this data-first approach?

Overcoming the challenges of a data-first approach

A data-first approach to machine learning involves some unique challenges, including management, analysis, and labeling.

Because machine learning requires a great deal of iteration and experimentation, companies often find themselves with a management system that’s a patchwork of models and results, stored haphazardly. Without a centralized spot for data storage and standard, reliable tools for exploration, results become difficult to track and reproduce, and finding patterns in the data becomes a challenge.

That means teams are often overwhelmed when digging out the insights they need from their data. Of course, large quantities of data is technically the way to solve business problems. But unless teams can streamline the data labeling process by labeling only the data that has true value, the process will quickly become unmanageable.

Using data to build a competitive advantage

Building an AI data engine is a series of iteration loops, with each loop making the model better. As companies with the best training data generally produce the most performant models, these companies will attract more customers who will generate even more data. It continuously imports model outputs as pre-labeled data, ensuring that each cycle is shorter than the last for labelers. That data is used to improve the next iteration of training and deployment, again and again. This ongoing loop keeps your models up to date, boosts their efficiency, and strengthens your AI.

Building this often required a great deal of hands-on labeling from subject matter experts — medical doctors identifying images of tumors; office workers labeling receipts; and so on. Automation dramatically speeds up the process, sending labeled data to humans to check and correct, eliminating the need to start from scratch.

A robust data engine needs only the smallest set of data to label to improve model performance, automatically labeling a sample of data for the model to work with, and only requiring verification from humans in some instances.

Putting it all together to improve model performance

Speeding up your data-centric iteration process takes just a few steps.

The first is to bring all your data to a single place, enabling your teams to access the training data, metadata, previous annotations, and model predictions quickly at any time, and iterate faster. Once your data is accessible within your training data platform, you can annotate a small dataset to get your model going.

Then, evaluate your baseline model. Measure your performance early, and measure it often. One or more baseline models can speed up your ability to pivot, as its performance develops. To create a solid foundation, your team should focus on identifying any errors early on and iterating, rather than optimizing.

Next, curate your data set according to your model diagnosis. Rather than bulk-labeling a massive amount of data, which takes time, energy, and money, create a small, carefully selected set of data to build on the baseline version of your model. Choose the assets that will best improve model performance, taking into account any edge cases and trends you found during model evaluation and diagnosis.

Finally, annotate your small dataset, and keep the iterative process going by assessing your progress and correcting for any errors like data distribution, concept clarity, class frequency errors, and outlier errors.

Training data platforms (TDP) are purpose-built for just this advantage, helping combine data, people, and processes into one seamless experience, and enabling ML teams to produce performant models quicker and more efficiently.

To learn more about boosting the performance of your model, reducing labeling costs, eliminating errors, solving for outliers and more, don’t miss this VB Live event!

Register here for free.

Attendees will learn how to:

  • Visualize model errors and better understand where performance is weak so you can more effectively guide training data efforts
  • Identify trends in model performance and quickly find edge cases in your data
  • Reduce costs by prioritizing data labeling efforts that will most dramatically improve model performance
  • Improve collaboration between domain experts, data scientists, and labelers


  • Matthew McAuley, Senior Data Scientist, Allstate
  • Manu Sharma, CEO & Cofounder, Labelbox
  • Kyle Wiggers (moderator), AI Staff Writer, VentureBeat

Repost: Original Source and Author Link


Xbox Live Games with Gold for November offers up two indie gems

Just as Sony has revealed the PlayStation Plus games for November, so too has Microsoft announced the next batch of titles that will be available through Xbox Games with Gold. As always, Xbox Live Gold subscribers will be getting four different games to download next month. Two of these games will be Xbox One titles, while the other two are Xbox 360 games playable through backward compatibility. All of the games offered through Games with Gold will be playable on Xbox Series X|S, even though none were made specifically for the platform.

The first game on offer will likely please fans of physics-based gameplay. Moving Out will be free for Xbox Series X|S and Xbox One owners, and as the name suggests, it’ll put players in the role of movers, tasking them with moving marked items without damaging them. The game can be played solo or in co-op with up to four players total, and we imagine that co-op play gets particularly zany.

Kingdom Two Crowns is the next game on tap for Xbox One and Xbox Series X players. Kingdom Two Crowns is a strategy game that blends tower defense and resource-management mechanics. Despite how complex that sounds, there isn’t a ton of hands-on gameplay, as players will mostly be riding their horse across the screen and collecting money to fund their kingdom’s defense. Kingdom Two Crowns may not be a big name, but it has some stellar user reviews on places like Steam, so it could be well worth the bandwidth to download it.

For the backward compatible games, we’ve got Rocket Knight and Lego Batman 2 DC Super Heroes, both for Xbox 360. While Lego Batman 2 likely needs no introduction, it’s worth highlighting Rocket Knight for anyone who grew up with the Sega Genesis. Rocket Knight, of course, was the main character in Sparkster and Rocket Knight Adventures, and he made his return to starring in video games with the Xbox 360.

As always, Microsoft is staggering these game releases throughout the month. Moving Out will be available November 1st-30th, while Kingdom Two Crowns will go free on November 16th and remain available through December 15th. Rocket Knight, meanwhile, will be available from November 1st-15th, and Lego Batman 2 DC Super Heroes will close the month out with availability from November 16th-30th.

Repost: Original Source and Author Link