Methodology
When you log food with a photo or by voice, Microgram estimates what you ate and how much. It then matches each food to an entry in the Nutrition Coordinating Center Food and Nutrient Database (NCCDB), and uses that entry's nutrition data to calculate calories, macros, and micronutrients for the estimated portion.
How does Microgram calculate nutrition from a photo?
Microgram uses computer vision to identify foods in a meal photo. It matches each food to a nutrition database entry and calculates calories and nutrients for the estimated portion.
How does Microgram estimate portion sizes?
For photo logs, Microgram uses visual cues, including depth, to estimate how much of each food is on the plate. For voice logs, Microgram uses your description, including any quantities you give, such as "half a cup of rice."
What does Microgram estimate, and what comes from a database?
For photo and voice logs, Microgram uses the photo or description to estimate what you ate and how much. When you scan a barcode or select a food through search, Microgram uses the database entry for the scanned product or selected food. With every logging method, Microgram calculates calories, macros, and micronutrients from the database values for the portion logged.
How accurate is Microgram's AI food logging?
Accuracy depends on what you're logging. Mixed meals like soups, stews, and smoothies can be harder for Microgram's AI to analyze from a photo because the ingredients are cooked or blended together. That makes it harder to tell what's in the meal and how much of each ingredient it contains.
You can easily edit the log to add, remove, or replace ingredients and adjust the amount of each one.