Honestly, this is probably the best link on PH today. The technology looks amazing, and on par with the type of ML that the big tech giants are working on. In terms of user scenario--this really hits the spot! Very well executed--it'll be great to see where the product heads to next, and what integrations are coming up (ie: FitBit, HealthKit, etc.)
Hey Product Hunt!
The @loseit team is excited to introduce Snap It, the first-ever, consumer-ready, food photo analysis and tracking feature to hit the market. Snap It uses some of today’s most advanced image recognition technology to deliver the best consumer food tracking experience in the world.
Snap It makes food tracking as easy as taking a picture. Simply upload a photo to Lose It! and Snap It will suggest foods it identifies, along with the calorie count and nutritional information, removing the #1 reason dieters quit tracking: the burden associated with logging meals.
How Snap It Beta works today:
* Open the Lose It! app and select a meal (breakfast, lunch, dinner or snack)
* Snap a picture of your meal
* Snap It will analyze your photo of the meal and provide food suggestions based on what it sees
* Users can then provide additional specifics for their photos. For example, if a user takes a photo of sushi, he or she can adjust the entry to indicate it was a salmon avocado roll or made with brown rice instead of white rice
* Select serving size
* Confirm and meal is logged
Snap It’s deep learning framework will learn and improve as users access the feature. Lose It!’s experience working with millions of users who track their food and exercise has given us unique access to data and insight that we can directly apply to make Snap It better. As these users begin using Snap It Beta, this dataset will grow dramatically and our ability to make useful, accurate and smart predictions from photos will continue to improve. Our goal is to evolve Snap It to a place where it can automatically recognize the specific foods, nutritional values and portion sizes in a photo.
Learn more at www.loseit.com/snapit
You made this look pretty & simple. I'd get away from the "losing weight" only angle & embrace the #QuantifiedSelf community... @loseit@dragonstyle@epuidokas@frank_denbow
Report
It isn't perfect yet but it is incredible
Report
Hi Eric, does it know what you're eating when you've made something froms scratch? For example, if I make pasta sauce from scratch, will it know what I have in there?
@mrabramowitz We can often identify sauce, but we're not to the point yet where we can break down ingredients further in that case.
With enough data, our model should start to learn what ingredients compose some foods, but there will always be cases where it's impossible from a photo alone. For example, we won't ever know exactly what's inside your burrito based on a photo from the outside- unless you take a bite before snapping the pic :)
Report
Haven't tried the app yet but I am very excited about this. I have always wanted to regularly track my meals but always stopped after a few days -- it was too annoying to manually enter in everything. I can imagine this will help a lot of people (like me!) track their meals and stick to diets
Report
Can I take a picture of the nutritional facts label on a wrapper?
Replies
Play DOTA2 for the ACLU!
Lose It! Calorie Counter
Fold
Molr Dental Club
INK'A
ArtSquare
Lose It! Calorie Counter