identif.Ai is most useful when calculating the nutritional values of plate of food.
It identifies each item on the plate, calculates the relative size and provides the most accurate macronutrient estimate.
The greatest value from this tool comes when it is not possible to weigh each food item, and when exact nutritional information is not available.
The AI model can only process what it can see. If any of the food items are fully or partially covered the results may not be as accurate as possible
Use the additional information text field to help the model when not all food is visible, or might be difficult to identify.
When taking a picture of the plate of food, try and provide some context to help to model - such as placing your hand next to the plate as a size reference.
Taking the picture directly from above may hinder the model by removing depth perception, try taking it at a slight angle.
The 'additional information' text field can be a very useful tool.
If the food could be difficult to identify (e.g. cream of mushroom mushroom spaghetti may look too similar to spaghetti carbonara), providing details to the model will help get the best results:
"The spaghetti is cream of mushroom spagetti, with cream, mushrooms and bacon"