Photo Diary & Macro Tracker
advertisement
| Naam | SnapCalorie |
|---|---|
| Versie | 120.0.0 |
| Update | 28 okt. 2024 |
| Grootte | 91 MB |
| Categorie | Gezondheid en fitness |
| Installaties | 50K+ |
| Ontwikkelaar | Perception Labs, Inc. |
| Android OS | Android 7.0+ |
| Google Play ID | com.snapcalorie.alpha002 |
SnapCalorie · Schermafbeeldingen
SnapCalorie · Beschrijving
Founded by ex-Google AI researchers, SnapCalorie’s photo calorie counter is the easiest and fastest way to track your diet and nutrition. Simply snap a photo or record a quick voice note to log an entire meal or day. Our AI algorithm will instantly provide you with calories, macros, and over 30 micronutrients based on verified USDA database values.
Do you only track calories?
No! We track all macronutrients and over 100 micronutrients. SnapCalorie's AI nutritionist can help you achieve a diverse set of dietary goals.
How accurate is it?
Our photo calorie counter is approximately twice as accurate as visually estimating portion sizes.
Our voice note algorithm is as accurate as the information you give it. Dictating gram values and ingredients as you place items on a food scale allows you to log your food with laboratory-grade precision and NO TYPING.
Don’t want to use a kitchen scale?
Our photo calorie counter can scan the exact volume of your food with the LiDAR depth sensor on your iPhone Pro or our voice note feature can assume average portion sizes if not specified.
How do you measure accuracy?
We were founded by a team of ex-Google AI researchers who co-founded Google Lens and Cloud Vision API. Our AI algorithm is the only one backed by peer-reviewed academic research. In our research, Nutrition5k, we collected a test dataset of 5,000 unique dishes, weighing every ingredient that went onto the plate. To evaluate our algorithm's accuracy, we ran the photo calorie counter on this dataset and compared the results to the true nutrient values.
Average expected error for a 500 calorie dish is +/- 80 calories on an iPhone Pro and +/- 130 on a regular iPhone. By comparison users eyeballing portion size visually were +/- 265 calories on average.
How does it work?
Snap a picture with the photo calorie counter and our AI starts by identifying the different types of food and where they are on the dish, like a nutritionist would. Next, if you have an iPhone Pro, we measure the food's volume with the LiDAR depth sensor. Our AI estimates portion size visually for phones without a depth sensor. Finally, we look up the nutritional values for that type of food and portion size in a trusted database (e.g., USDA) and sum up the totals for you!
If anything looks incorrect, you can fix it yourself OR send it back to our team of nutrition experts for review.
Over time, our photo calorie counter fine-tunes itself based on your specific data, learning from the corrections you make. Always eat meat alternative burgers? Our AI algorithm will slowly learn and adapt to you.
What about cooking fats and oils?
Cooking fats are challenging for anyone counting calories, and our algorithm is no exception. You can add voice notes to hint at the cooking method used or just adjust cooking fats after the initial prediction.
What if something is missing from your barcode database?
One of our favorite features is the ability to take a photo of a nutrition label. Our AI will type in all the values for you in seconds!
Terms: https://www.snapcalorie.com/terms/
Do you only track calories?
No! We track all macronutrients and over 100 micronutrients. SnapCalorie's AI nutritionist can help you achieve a diverse set of dietary goals.
How accurate is it?
Our photo calorie counter is approximately twice as accurate as visually estimating portion sizes.
Our voice note algorithm is as accurate as the information you give it. Dictating gram values and ingredients as you place items on a food scale allows you to log your food with laboratory-grade precision and NO TYPING.
Don’t want to use a kitchen scale?
Our photo calorie counter can scan the exact volume of your food with the LiDAR depth sensor on your iPhone Pro or our voice note feature can assume average portion sizes if not specified.
How do you measure accuracy?
We were founded by a team of ex-Google AI researchers who co-founded Google Lens and Cloud Vision API. Our AI algorithm is the only one backed by peer-reviewed academic research. In our research, Nutrition5k, we collected a test dataset of 5,000 unique dishes, weighing every ingredient that went onto the plate. To evaluate our algorithm's accuracy, we ran the photo calorie counter on this dataset and compared the results to the true nutrient values.
Average expected error for a 500 calorie dish is +/- 80 calories on an iPhone Pro and +/- 130 on a regular iPhone. By comparison users eyeballing portion size visually were +/- 265 calories on average.
How does it work?
Snap a picture with the photo calorie counter and our AI starts by identifying the different types of food and where they are on the dish, like a nutritionist would. Next, if you have an iPhone Pro, we measure the food's volume with the LiDAR depth sensor. Our AI estimates portion size visually for phones without a depth sensor. Finally, we look up the nutritional values for that type of food and portion size in a trusted database (e.g., USDA) and sum up the totals for you!
If anything looks incorrect, you can fix it yourself OR send it back to our team of nutrition experts for review.
Over time, our photo calorie counter fine-tunes itself based on your specific data, learning from the corrections you make. Always eat meat alternative burgers? Our AI algorithm will slowly learn and adapt to you.
What about cooking fats and oils?
Cooking fats are challenging for anyone counting calories, and our algorithm is no exception. You can add voice notes to hint at the cooking method used or just adjust cooking fats after the initial prediction.
What if something is missing from your barcode database?
One of our favorite features is the ability to take a photo of a nutrition label. Our AI will type in all the values for you in seconds!
Terms: https://www.snapcalorie.com/terms/
