Fitness AI: How synthetic data powers better workouts

Were you unable to attend Transform 2022? Check out all of the summit sessions in our on-demand library now! Watch here.

Can AI-driven fitness apps, developed with synthetic data, pump up your workout? 

During the COVID-19 pandemic, home fitness apps were all the rage. From January through November 2020, approximately 2.5 billion health and fitness apps were downloaded worldwide. That trend held and shows no signs of slowing down, with new data predicting growth from $10 million in 2022 to $23 million by 2026. 

As more people use fitness apps to train and track their development and performance, fitness apps are increasingly using AI to power their offerings by providing AI-based workout analysis, incorporating technologies including computer vision, human pose estimation, and natural language processing techniques.

Tel-Aviv-based Datagen, which was founded in 2018, claims to provide “high-performance synthetic data, with a focus on data for human-centric computer vision applications.”

The company just announced a new domain, Smart Fitness, on its self-service, visual synthetic data platform that helps AI developers produce the data they need to analyze people exercising and train smart fitness equipment to “see.” 

“At Datagen, our focus is to aid computer vision teams and accelerate their development of human-centric computer vision tasks,” Ofir Zuk, CEO of Datagen, told VentureBeat. “Almost every use case we see in the AI space is human-related. We are specifically trying to solve and help understand the interconnection between humans and their interaction with surrounding environments. We call it human in context.” 

Synthetic visual data represents fitness environments

The Smart Fitness platform provides 3D-annotated synthetic visual data in the form of video and images. This visual data accurately represents fitness environments, advanced motion, and human-object interactions for tasks related to body key point estimation, pose analysis, posture analysis, repetition counting, object identification and more. 

In addition, teams can use the solution to generate full-body in-motion data to iterate on their model and improve its performance quickly. For example, in cases of pose estimation analysis, an advantage the Smart Fitness platform provides is the capability to quickly simulate different camera types for capturing a variety of differentiated exercise synthetic data.

image gif showing datagens body key point analysis for workouts
Source: Datagen

Challenges to training AI for fitness

Pose estimation, which is a computer vision technique that helps determine the position and orientation of the human body with an image of a person, is one of the unique solutions that AI has to offer. It can be used in avatar animation for artificial reality, for example, as well as markerless motion capture and worker pose analysis. 

To correctly analyze posture, it is necessary to capture several images of the human actor with its interacting environment. A trained convolutional neural network then processes these images to predict where the human actor’s joints are located in the image. AI-based fitness apps generally use the device’s camera, recording videos up to 720p and 60fps to capture more frames during exercise performance. 

The problem is, computer vision engineers need vast amounts of visual data to train AI for fitness analysis when using a technique like pose estimation. Data involving humans performing exercises in various forms and interacting with multiple objects is highly complex. The data must also be high-variance and sufficiently diverse to avoid bias. Collecting accurate data which covers such a variety is nearly impossible. On top of that, manual annotation is slow, prone to human error, and expensive. 

While an acceptable level of accuracy in 2D pose estimation has already been reached, 3D pose estimation lacks in terms of generating accurate model data. That is especially true for inference from a single image and with no depth information. Some methods make use of multiple cameras pointed at the person, capturing information from depth sensors to achieve better predictions. 

However, part of the problem with 3D pose estimation is the lack of large annotated datasets of people in open environments. For example, large datasets for 3D pose estimation such as Human3.6M were captured entirely indoors to eliminate visual noise.

There is an ongoing effort to create new datasets with more diverse data regarding environmental conditions, clothing variety, strong articulations, and other influential factors.

The synthetic data solution

To overcome such problems, the tech industry is now widely using synthetic data, a type of data produced artificially that can closely mimic operational or production data, for training and testing artificial intelligence systems. Synthetic data offers several significant benefits: It minimizes the constraints associated with the use of regulated or sensitive data; can be used to customize data to match conditions that real data does not allow; and it allows for large training datasets without requiring manual labeling of data.

According to a report by Datagen, the use of synthetic data reduces time-to-production, eliminates privacy concerns, provides reduced bias, annotation and labeling errors, and improves predictive modeling. Another advantage of synthetic data is the ability to easily simulate different camera types while generating data for use cases such as pose estimation. 

Exercise demonstration made simple

With Datagen’s smart fitness platform, organizations can create tens of thousands of unique identities performing a variety of exercises in different environments and conditions – in a fraction of the time. 

“With the prowess of synthetic data, teams can generate all the data they need with specific parameters in a matter of a few hours,” Zuk said. “This not only helps retrain the network and machine learning model, but also allows you to get it fine-tuned in no time.”

datagen's Smart Fitness platform dashboard
Source: Datagen

In addition, he explained, the Smart Fitness platform optimizes your ability to capture millions of substantial visual exercise data, eliminating the repetitive burden of capturing each element in person. 

“Through our constantly updating library of virtual human identities and exercise types, we provide detailed pose information, such as locations of the joints and bones in the body, that can help analyze intricate details to enhance AI systems,” he said. “Adding such visual capabilities to fitness apps and devices can significantly improve the way we see fitness, enabling organizations to provide better services both in person and online.”

data gen Smart Fitness platform post tracking gif of person doing lunge
Source: Datagen

Fitness AI and synthetic data in the enterprise

According to Arun Chandrasekaran, distinguished VP Analyst at Gartner, synthetic data is, so far, an “emerging technology with a low degree of enterprise adoption.” 

However, he says it will see growing adoption for use cases for which data must be guaranteed to be anonymous or privacy must be preserved (such as medical data); augmentation of real data, especially where costs of data collection are high; where there is a need to balance class distribution within existing training data (such as with population data), and emerging AI use cases for which limited real data is available. 

Several of these use cases are key for Datagen’s value proposition. When it comes to enhancing the capabilities of smart fitness devices or apps, “of particular interest will be the ability to boost data quality, cover the wide gamut of scenarios and privacy preservation during the ML training phase,” he said. 

Zuk admits that it is still early days for bringing AI and synthetic data, and even digital technologies overall, into the fitness space. 

“They are very non-reactive, very lean in terms of their capabilities,” he said. “I would say that adding these visual capabilities to these fitness apps, especially as people exercise more in their own home, will definitely improve things significantly. We clearly see an increase in demand and we can just imagine what people can do with our data.” 

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

Repost: Original Source and Author Link


Pikmin Bloom is Basically a Fitness App, But That’s Ok

Pikmin Bloom recently released for mobile devices across the world. Nintendo fans weren’t sure what to expect when Pokémon Go developer Niantic announced a new augmented reality game based on Pikmin, the popular franchise starring half-plant, half-animal creatures. While the game takes some inspiration from Pokémon Go, it’s a completely different experience that’s more akin to a fitness app than an actual game.

Niantic calls Pikmin Bloom “an app that brings a little joy to your everyday journeys on foot.” This checks out, considering it seems like more of an activity tracker skinned as a Pikmin game. It features walking as a core gameplay element, and progress heavily relies on how many steps you take and where you go. Players level up based on the number of steps they take, how many Pikmin they grow, and how many expeditions they complete. They can also unlock Decor Pikmin, or Pikmin dressed in costumes, as collectibles.

Pikmin Bloom also asks players to plant flowers to build a more lush neighborhood. Players feed Pikmins nectar, which in turn grows the leaves on their heads into glowing flowers. Tapping on these flowers turns them into petals that the players can plant around their neighborhood as they walk. Players can also find expeditions to send their Pikmin on throughout the area, which can yield even more nectar and seedlings to grow their team. At approximately 9 p.m. local time, the app compiles your daily progress into a log where you can write a little journal entry for yourself.

It’s a way fluffier experience than the original game. In Pikmin, Captain Olimar employs the help of the titular plantlike creatures to help him gather parts for his spaceship so that he can return home. However, there’s no need to throw Pikmin into dangerous territories with deep puddles or toxic fumes in Pikmin Bloom. Players don’t have to worry about monstrous predators eating their plant/alien friends. They also don’t need to worry about sending a player character in a perilous situation back home to their family.

Unfortunately, that also means Pikmin Bloom lacks stakes. It also doesn’t provide as much engagement as Pokémon Go, where players can find Pokémon at random spawn points throughout their area. Pikmin Bloom doesn’t have as many of these points of interest. So, for anyone living in a suburban area (like me), the map likely looks like an empty green field without anything to explore. You might come across some sprouts that you can help bloom into flowers around your area and trails of flowers left from walks, but that’s it.

Pikmin squad on screen

Pikmin Bloom mainly encourages activity through its gameplay. Growing a team of Pikmin means pulling seedlings when they’re ready — approximately every 1,000 steps. Leveling up also requires a certain number of steps, which in turn unlocks more Pikmin and other in-game features. Players need to reach level six before they unlock expeditions, one of the main ways to gather nectar and seedlings in the game. Thankfully, the game lists what you need to do to level up on your profile.

Pikmin Bloom lacks purpose as a “game,” even if it offers something to do during daily walks. I wouldn’t recommend it to people who don’t like walking because then you basically won’t be able to play. It might be an appealing experience to those who have played Pikmin, but I can’t imagine it having as much mainstream appeal as Niantic’s other games. It effectively pushed me into walking to get more Pikmin, but there’s not much besides curiosity and nostalgia keeping it on my phone.

Pikmin Bloom is now available on iOS and Android devices. It’s free, so there’s no downside to trying it next time you’re out and about.

Editors’ Choice

Repost: Original Source and Author Link

Tech News

Buying a smart toy or fitness tracker? Research how safe they are first

The wearable technology market is booming, with half a billion wearables sold globally in 2020. Apps on these devices, or the devices themselves, often claim to monitor our health to spot illnesses, track our workouts to help us reach our fitness goals, or keep an eye on our children’s whereabouts to enhance their safety.

But they’re also divisive. Supporters of wearable technology claim that health trackers should be prescribed by the NHS and could even deliver an early warning of a possible COVID-19 infection. GPS tracking devices designed to be worn by children, meanwhile, are seen as a safety asset for parents.

Yet studies have found fitness trackers to be too inaccurate and misleading to be used by medical professionals, and that, because they’ve been rushed to market, wearables of all kinds are an insecure “Wild West” region of technology that requires urgent regulation.

In a recent report, we looked at the security risks associated with wearable devices, as well as “smart toys” that can record children in their homes. We found a concerning lack of security – especially for devices aimed at children – which lack even the most basic cybersecurity precautions, leaving them open to abuse.

Fitness trackers and personal data

One key issue with wearables is the data they generate and share. For instance, many fitness trackers rely on data on a person’s location to map their workouts. That’s great if you’re keen to track the distance of your jogs, but it’s not especially sensible if you’re embarking on those jogs from a military base in hostile territory.

Beyond that specific example, which caused some embarrassment for the US military in 2018, it’s clear that sharing your location publicly, even in a safe civilian setting, comes with significant risks.

And it’s not just the real-time tracking of your running route that could expose your whereabouts. Because these trackers upload your workouts to an app and share them publicly, it’s possible for predators to use historic running, biking or hiking routes to predict where you might be at a given time. This safety issue isn’t only restricted to workouts. Even something as innocuous as sharing a photo through your Apple watch can give away your geolocation.

Are trackers safe for children?

Even more concerning are devices designed to be worn by children, sales of which are expected to reach $875 million (£620 million) by 2025. These watches are marketed as wearable tech to keep kids safe, tracking their location and alerting parents when the watch’s onboard “SOS” button is pressed – or if the child travels beyond a geofenced area.

Smart watches as safety devices on children’s wrists may sound like a boon for anxious parents, but a 2017 survey of children’s smart watches found that the all-important “SOS” button either got stuck or didn’t work at all in most cases.

Additionally, flaws in some smart watches’ accompanying apps have raised serious safety concerns. Security researchers have found they could not only easily access children’s historical route data – like their path to and from school – and monitor their geolocation in real time, but they could also speak directly to the child, through the watch, without the call being reported in the parent’s app.

Connected toys

Fears that internet of things devices can give people unauthorised access to children also extend to the “smart toy” market. Some of these toys contain hidden cameras and microphones which, if hacked, could be used to record the interior of your home, including children’s rooms.

In 2017, German regulators recognised this danger by banning the sale of the Cayla “smart doll”, labelling it as the kind of “de facto espionage device” that Germany’s Telecommunications Act legislates against. In an unusual and unsettling move, the regulator went further by asking parents who’d bought one to destroy the doll to prevent illicit surveillance.

Even if the manufacturers of smart toys and children’s smart watches can guarantee far better security than that which led to the Cayla ban, there remain other surveillance concerns. In 2019, a UNICEF-led report highlighted how children’s rights – to creativity, freedom of choice and self-determination – are challenged by smart devices. Present in schools, at home, and on the wrist, this kind of round-the-clock surveillance, the report argues, restricts carefree childhood and hurts kids’ development.

Making trackers safer

Trackers and toys can be made safer. Before we allow these devices to flood the market, it’s essential we standardise the minimum security requirements that manufacturers must comply with – no matter where in the world these devices are made.

Key among these standards should be the removal of factory-default passwords on devices – which, like “admin” or “1234”, are easily guessed or discovered by even the most novice hacker. Manufacturers should also publish a vulnerability disclosure to help users understand risks, and make regular software updates in response to vulnerabilities unearthed by security researchers.

Clearly, monitoring people’s health via wearable trackers has the potential to radically improve access to medical care. Likewise, every parent wants their child to be safe, and smart devices, like mobile phones before them, could be a reliable tool for checking in with them. But without safety standards, these devices have the potential to cause more harm than they offset. Regulators must act fast to stop this growing market from leading to significant harms.

Article by Saheli Datta Burton, Research Fellow, Department of Science Technology Engineering and Public Policy, UCL and Madeline Carr, Professor of Global Politics and Cybersecurity, UCL

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Did you know we have a newsletter all about consumer tech? It’s called Plugged In –
and you can subscribe to it right here.

Repost: Original Source and Author Link

Tech News

Amazon’s Halo fitness wearable wants to watch you move

Amazon’s Halo fitness band will soon be able to assess how you move, with a new – and somewhat confusing – tracking feature for the controversial wearable. Launched last year, Halo aims to bring AI and machine learning to bear on individual fitness, though concerns around just what data was being shared with Amazon as part of that led to a troubled debut.

Rather than just logging exercise, steps, and other metrics like other fitness wearables do, Halo also promised to make individual body assessments. To do that, users were asked to take photos of themselves – while wearing tight underwear or workout clothes – that were uploaded to Amazon’s servers. Its machine learning then figured out things like body fat percentage and other metrics.

It was a clever system, but unsurprisingly ran into discomfort about putting uploading such images online – even if Amazon promised to delete them once the analysis had been made. Now, this new Movement Health feature for Halo once again wants to use your smartphone’s camera to share with the AI, only this time Amazon is asking for video too.

Movement health, Amazon says, is based on three factors: posture, stability, and mobility. It’s basically our ability to undertake everyday motions – like carrying items, reaching for things on high shelves, or running – along with our tendencies to slump while sitting at a computer or lean one way or another while standing. That, Amazon argues, is just as important as being able to run 5 miles.

Halo’s answer starts with a video assessment, your smartphone recording you doing five movements: single leg balances, forward lunges, overhead squats, overhead reaches, and feet-together squats. It then calculates body position and any issues with the three movement categories, and boils all that down to a Movement score out of 100. You’ll also get a report on your performance in those categories, plus details on four key areas of the body: the trunk, hips, lower body, and shoulders.

After that, Halo will serve up “corrective exercise videos” focused on your individual trouble spots. That could include stretches, balances, or breathing exercises; a set takes 5-10 minutes, Amazon says, and it recommends they be completed at least three times a week. Repeat assessments every 2-4 weeks track progress, and there’ll be more workout recommendations for an exercise-focused add-on if you want to include calorific burn as well.

Obviously the efficacy of all this depends on a number of factors, not least Amazon’s core model of Movement Health itself. The accuracy of that, and what you’re being compared to, will clearly have an impact on how realistic your score out of 100 is. There are also legitimate questions around how useful, or reasonable, expressing something so broad as movement as a numerical score actually can be.

Indeed, the privacy issue is arguably the least of the concerns here. Amazon says that its movement assessment videos are encrypted as they’re uploaded, “processed within seconds,” and then promptly deleted. Nobody sees them, and you can’t even keep copies yourself for later review.

Movement Health will be added to Amazon Halo in the coming weeks, the company says. The Halo wearable itself is $99.99, including six months of membership; after that it’s $3.99 per month, though the band does offer basic tracking with no subscription fee.

Disclosure: SlashGear uses affiliate links, If you click on a link in this article and buy something we’ll get a small cut of the sale.

Repost: Original Source and Author Link

Tech News

What’s VO2Max and what does it tell you about your fitness?

Standards is a series explaining various tech protocols, looking at their practical implementation, and where we could find them in devices around us.

Fitness bands and watches have thrown around numbers at our faces and motivate us to be more healthy.

All these smart wearables have two primary measures: the number of steps taken and the number of active calories burnt in a day.

Apple pioneered the system to display three rings for calories, steps, and standing. It edges you to complete them out for the day to feel we’ve been active enough. Other smart band makers have also followed rings or a similar system for daily fitness goals.

However, another fitness stat is gaining prominence in these wearable devices: VO2Max. It’s a metric that allows devices to calculate the intensity of your workout and cardiovascular strength.

What is VO2Max?

VO2Max stands for the maximum volume of O2 (oxygen) you can utilize during a workout. The measure’s unit is ml/kg/min, indicating that it takes into account the amount of oxygen in ml you can utilize per kg of body weight every minute. The score helps you improve your performance in activities that require better cardio fitness.

What’s it used for?

VO2Max is a measure of your cardiovascular fitness or endurance in simple words. It can give you a sense of how well you utilize oxygen while doing physically strenuous activity. The higher your VO2Max score, the more likely you can perform while playing sports, running, or cycling.

Unlike step counts and calories, VO2Max is not a universal score. It depends on your age, the work you’re putting in, and your activity over time.

Credit: Garmin/Cooper Institute