Time to throw away your scales?

mirrim52

I wouldn't worry about it. It is just a research project. No plans for release. But it does ask you to verify what it sees, apparently.

From http://techxplore.com/news/2015-06-google-im2calories-calories-meal-photo.html

"With Im2Calories, for example after a picture is taken and the system analyzes what it sees, it spits out a report listing what it believes it has found: poached eggs, buttered toast, two strips of bacon, etc. At that point, the user has the option to make corrections, telling the system that in fact, the eggs were boiled, or fried, or whatever. The system would then make that correction, and add a note to its database so that it would not make the same mistake in the future.

Presumably, such an app would undergo some initial training before being sent out into the populace, though in this case it is not clear if that will happen at all with Im2Calories. Murphy said the app was being developed as a research project, which means that at this time there are no plans to actually release it. That could change of course if the app proves highly accurate and if users hear of it and beg for Google to give it to them."

Sued0nim

Chrysalid2014 wrote: »

rabbitjb wrote: »

TimothyFish wrote: »

I agree with the developers. Even if they are off by 20%, I'm still impressed.

Now I get how you follow the programme and I agree that portion estimating can work if it's how you work it

But how on earth can you agree when two food items can look exactly the same but have thousands of calories in difference?

How can any visual recognition software work out ingredients from a cooked dish

Is ridiculous to imagine a 20% margin of error

Perhaps it all levels out in the long run... i.e. sometimes the calorie estimate will be 20% over and sometimes 20% under.

I think there are good visual clues as to how something has been cooked. My chicken looks a lot different if I fry it with/without oil, for example.

And if it was something like a stew, perhaps it analyses the section of the photo it can see, identifies the ingredients (type and ratio) and then assumes the same ratio for the entire quantity.

It would be interesting to test this thing.

We are in the realm of Isaac Asimov here

Alluminati

Yeah this is a fabulous idea.
Just kidding.

wizzybeth

PaulaWallaDingDong wrote: »

Here's some more riduclousness, in case just the concept of it didn't seem silly enough... http://www.theverge.com/2015/6/2/8707851/google-calories-food-photos-im2calories

That article made me mad. Said calorie counting doesn't work. Obviously written by an idiot.

booksandchocolate12

mirrim52 wrote: »

I wouldn't worry about it. It is just a research project. No plans for release. But it does ask you to verify what it sees, apparently.

From http://techxplore.com/news/2015-06-google-im2calories-calories-meal-photo.html

"With Im2Calories, for example after a picture is taken and the system analyzes what it sees, it spits out a report listing what it believes it has found: poached eggs, buttered toast, two strips of bacon, etc. At that point, the user has the option to make corrections, telling the system that in fact, the eggs were boiled, or fried, or whatever. The system would then make that correction, and add a note to its database so that it would not make the same mistake in the future.

It sounds like it would be a lot more work than just logging what you eat.

BWBTrish

20% wrong over 1500 calories = 300 can be you whole deficit!

FitForL1fe

doesn't this whole idea seem like way more of a pain in the *kitten* than just setting something on top of a food scale and reading the numbers it displays?

BWBTrish

draznyth wrote: »

doesn't this whole idea seem like way more of a pain in the *kitten* than just setting something on top of a food scale and reading the numbers it displays?

lol i was thinking that too.
The only thing i can think of where it might be handy is when you eat out. You get a 20% of guesstimation of your food.

crazyjerseygirl

I think weighing/measuring is far more accurate BUT I think having photos analyzed is better than nothing.

Calorie counting forever is a difficult prospect, this might be a good maintenance tool when coupled with regular weighing of oneself, going stricter if you see a gain.

It might also be a good starting point for those who think tracking is too difficult, 20% is high inaccuracy, but if you see that youre eating roughly 4000kcal a day it can be a real eye opener.

So conclusion, not for me now, but after I've lost the weight and learned my mindfulness I might be willing to give it a go as a maintenance tool.

ETA: it might also be useful for restaurant food. I'm not bringing a scale to a steakhouse but if I could get a rough estimate I could use that along with a calorie counting app

FitForL1fe

crazyjerseygirl wrote: »

I think weighing/measuring is far more accurate BUT I think having photos analyzed is better than nothing.

Calorie counting forever is a difficult prospect, this might be a good maintenance tool when coupled with regular weighing of oneself, going stricter if you see a gain.

It might also be a good starting point for those who think tracking is too difficult, 20% is high inaccuracy, but if you see that your eating roughly 4000kcal a day it can be a real eye opener.

So conclusion, not for me now, but after I've lost the weight and learned my mindfulness I might be willing to give it a go as a maintenance tool.

I'd bet the nutrition info that fast food restaurants are required to provide these days has a smaller margin of error than 20%

crazyjerseygirl

PaulaWallaDingDong wrote: »

Here's some more riduclousness, in case just the concept of it didn't seem silly enough... http://www.theverge.com/2015/6/2/8707851/google-calories-food-photos-im2calories

That whole article is just anti-CICO, not really anti-Google app

crazyjerseygirl

draznyth wrote: »

crazyjerseygirl wrote: »

I think weighing/measuring is far more accurate BUT I think having photos analyzed is better than nothing.

Calorie counting forever is a difficult prospect, this might be a good maintenance tool when coupled with regular weighing of oneself, going stricter if you see a gain.

It might also be a good starting point for those who think tracking is too difficult, 20% is high inaccuracy, but if you see that your eating roughly 4000kcal a day it can be a real eye opener.

So conclusion, not for me now, but after I've lost the weight and learned my mindfulness I might be willing to give it a go as a maintenance tool.

I'd bet the nutrition info that fast food restaurants are required to provide these days has a smaller margin of error than 20%

I can't really speak to that. I do wish that the 20% was more accurately defined though. Is that 20% per day or per food item, does the app tend to over or under estimate, or is it a draw. If the 20% is an average is that app better at estimating some foods better than others or is it all clustering around 20%?

FitForL1fe

crazyjerseygirl wrote: »

draznyth wrote: »

crazyjerseygirl wrote: »

I think weighing/measuring is far more accurate BUT I think having photos analyzed is better than nothing.

Calorie counting forever is a difficult prospect, this might be a good maintenance tool when coupled with regular weighing of oneself, going stricter if you see a gain.

It might also be a good starting point for those who think tracking is too difficult, 20% is high inaccuracy, but if you see that your eating roughly 4000kcal a day it can be a real eye opener.

So conclusion, not for me now, but after I've lost the weight and learned my mindfulness I might be willing to give it a go as a maintenance tool.

I'd bet the nutrition info that fast food restaurants are required to provide these days has a smaller margin of error than 20%

I can't really speak to that. I do wish that the 20% was more accurately defined though. Is that 20% per day or per food item, does the app tend to over or under estimate, or is it a draw. If the 20% is an average is that app better at estimating some foods better than others or is it all clustering around 20%?

I assume it's a margin of error

a 20% margin of error per item would be the same as 20% for the entire day

calories measured = actual calories +/- 20%. so between 0.8 and 1.2

rjmudlax13

Google (aka Skynet) is becoming self aware!