How an Indian startup is using smartphone’s front camera to make fitness fun

Two years in the past, when Anurag Mundhada and Jayesh Hannurkar, graduates from IIT-Bombay, have been fighting weight points in the course of the first wave of the coronavirus pandemic, they considered bringing their experience in sensors and pc imaginative and prescient to create an interactive house exercise platform.

“The largest drawback assertion from a shopper standpoint that we’re engaged on to resolve is motivation,” Mundhada defined the concept behind their Insane AI app. “Super match fashions and trainers can generally actually flip off novices however for those who make the preliminary onboarding easy and frictionless as potential, health could be enjoyable and is for everybody,” he stated.

Insane AI, which is at the moment in beta for effectively over a 12 months and nears official launch quickly, is completely different from different health apps, declare the founders. Unlike Apple Fitness Plus or Nike Training Club, Insane AI makes use of your smartphone’s entrance digital camera to trace your movement throughout a exercise, they stated.

“The digital camera tracks the physique and divides it into 21 factors: your shoulders, elbows, ankles, and knees. Once the monitoring of factors is finished in real-time, we run a video AI layer on high of that which analyses the movement and clarifies which train is being accomplished, counts the reps, after which the system offers a rating of how intense the train is,” Mundhada instructed indianexpress.com throughout a video name.

Anurag Mundhada (left) and Jayesh Hannurkar (proper) co-founded Insane AI in 2020.

“If you might be doing a squat and your hips are under your knees, you get a rating of 1 and in case you are simply doing very shallow reps, you’ll not get the identical quantity of rating,” he added. Mundhada claimed its AI-powered exercise monitoring know-how is 98 per cent correct based mostly on the corporate’s inside take a look at information.

The intention to make use of a smartphone as an integral a part of your health journey and create a brand new person expertise round cellphone cameras comes from Mundhada and Jayesh’s earlier expertise in digital camera sensing applied sciences. Jayesh was with Sony Japan earlier than co-founding Insane AI and labored on LiDAR sensors, a know-how that provides depth scanning for higher images and AR in smartphones and is predominantly present in Apple’s high-end iPhones and iPad Pros since 2020. Mundhada, then again, labored with a Bengaluru-based Axogyan AI the place he did a venture with Mercedes Benz for human sensing for autonomous autos.

Mundhada stated he and his crew have been fascinated by Nintendo Wii and Microsoft Kinect and wished to deliver the power to trace physique movement utilizing a cellphone digital camera. Although smartphones over time have gotten higher {hardware}, depth sensors nonetheless are restricted to a handful of high-priced telephones. “Using AI we enabled the traditional RGB digital camera to trace real-time exercise monitoring which is current on all telephones to do what Nintendo Wii or Xbox Kinect can do,” he stated when requested how a smartphone digital camera can monitor physique movement with no further sensors.

The Insane AI app makes use of your smartphone’s front-facing digital camera to trace your physique movement.

To take a look at whether or not the smartphone digital camera actually tracks physique movement and whether or not an app could be created round house exercises, Mundhada developed Pushup dojo which went reside within the early days of the pandemic. The app was used as a proof of idea to reveal that know-how and exercise monitoring utilizing a cellphone digital camera may very well be potential. It was a primary exercise app that garnered 2000 customers, largely mates and fast household of the co-founders. That gave Mundhada some confidence to transform the algorithms and person interface and ideal the physique movement monitoring know-how to create the Insane AI app in early 2021. This time, the crew labored with health trainers to create the app and supposed a large person base.

Mundhada calls Insane AI a house exercise app with a “extremely gamified expertise.” The app, which could be downloaded from the Google Play Store and Apple App Store, at the moment requires individuals to sign-up for the waitlist, and starting the final week of October, Insane AI will change into accessible to those that are at the moment on the waitlist.

To use the app, one has to first arrange a cartoon(y) avatar which can be your digital identification on the exercise platform. You can simply place the cellphone in entrance of you at a distance, begin exercising, and the digital camera then begins monitoring your physique movement in the course of the exercise. Every exercise or train customers do can be counted and they’re going to get rewarded via cash which could be redeemed within the type of digital collectibles and coupons for which the startup has tied up with a couple of companions. Users may problem their mates each in real-time in addition to turn-based health battles. The Insane AI app works on any smartphone that has been launched in 2016 and after. In the approaching days, customers will see a brand-new interface and a number of options together with much more exercises and workout routines.

Mundhada calls Insane AI a house exercise app with a “extremely gamified expertise.”

The AI-powered exercise know-how mannequin developed by Mundhada and his crew not simply tracks physique factors but in addition tracks customers’ health ranges throughout 5 completely different dimensions comparable to power, cardio, higher physique, and decrease physique. Mundhada stated they’ve taken lots of enter from health specialists to develop the exercise plans and determine on completely different health ranges, whereas the corporate’s in-house information science crew has labored on making a health mannequin which could be customised and personalised as per a person. The insane AI app gives 35 completely different workout routines and 150 exercises, out of which 4 exercises are free, to start with.


According to Mundhada, Insane AI appears to be extensively used at 7:30 am and one other peak comes within the night, round 6 pm. On common, customers spend 9 minutes on the app, up from 5 minutes within the early days of Insane AI. However, energy customers spend virtually an hour on the platform.

“Insane AI just isn’t meant to exchange a private coach fully,” Mundhada stated, including the app is “designed to have a simple pathway to studying health and getting extra superior.” “For circumstances comparable to excessive weight problems, it’s all the time advisable to have a human within the loop and that’s not only for studying but in addition for the prevention of accidents somebody who may help you notice you once you’re lifting weights.” Mundhada is evident that Insane AI goals to supply a rewarding and enjoyable expertise and complement your journey alongside the best way.

The house exercise app comes with a promise of not accumulating customers’ footage and movies however solely accumulating anonymised information of movement itself to enhance the AI engine which does the movement monitoring.

For monetisation, Mundhada says they’re going for a freemium mannequin that can let customers expertise the app free of charge, with sure options comparable to superior challenges and redeeming rewards requiring a subscription. Right now, those that sign-up for the app will stand up to 6 months of free subscription. Mundhada and the crew are but to determine what could be the price of a subscription.

Last 12 months, the start-up raised $900,000 in seed funding. The spherical was led by early-stage enterprise capital fund pi Ventures, Anupam Mittal, Sameer Pitalwalla, and different angels in sports activities, health, and tech. In the long term, Mundhada needs to show Insane AI right into a health metaverse and work on constructing a basis for an immersive world that has already been began.



https://news.google.com/__i/rss/rd/articles/CBMilgFodHRwczovL2luZGlhbmV4cHJlc3MuY29tL2FydGljbGUvdGVjaG5vbG9neS90ZWNoLW5ld3MtdGVjaG5vbG9neS9ob3ctYW4taW5kaWFuLXN0YXJ0dXAtaXMtdXNpbmctc21hcnRwaG9uZXMtZnJvbnQtY2FtZXJhLXRvLW1ha2UtZml0bmVzcy1mdW4tODIyNTgzNi_SAQA?oc=5

Recommended For You

About the Author: Adrian

Leave a Reply