HungryBot
AI
food image recognition
SHOW HN2016
202points
127comments
#134of 2,264 in 2016
Top 6%of 2016 by points
How HN took it
CuriousMostly questions and neutral comments
11 positive, 32 neutral or asking, 3 mixed, 10 critical, out of 56 comments about the project. Thread score +0.04.
Execution: 3 for, 3 againstIdea: 3 for, 2 againstPerformance: 1 for, 3 againstDesign: 0 for, 2 againstDocs: 1 for, 0 against
The thread, comment by comment
Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.
positive neutral or mixed critical about another commenter or something else the maker
What HN said
What people liked
Classification accuracy is impressive
From 2 comments
The classification looks good, congrats!. Where did you get the million labeled images to train your classifier?
npalli on HNBuilt using those newfangled convolutional neural network things. Trained caffe on a set of about a million food related images. Can't believe how accurate the classifier is...…
subcosmos on HN
Detection is great
From 1 comment
Your app kinda sucks now, but the detection is great. You could also do food identification as a service... I'm trying to use fatsecret/myfitnesspal for counting calories.…
swah on HN
Objections
AI accuracy is doubtful without a large database and users
From 1 comment
I would pay something like $5 per day for this if it was a real person who looks at my photos, categorizes them accurately, and enters them into MyFitnessPal. I won't trust an AI…
nathan_f77 on HN
Calorie estimation is hard due to hidden fats and cooking methods
From 3 comments
I agree with your suggestions for human assessment because of problem posed by food calories varying significantly due to different preparations. Take for example soda, if I gave…
HamSession on HNIt would be difficult to account properly for ingredients such as butter and oil which are hard to see in a finished dish and drastically affect the calories. But people would…
logn on HNImpossible. How can someone tell whether you cooked with butter and oil, for instance?
a_small_island on HN
Login via Facebook is unwelcome
From 2 comments
Login via Facebook? Come on! See you next app.
imacomputer2 on HNPretty nifty, but please consider allowing creation of an account with a simple email address, or even anonymously. I really don't want to have to connect my Facebook or other…
zeveb on HN
Incompatible with some Android devices
From 1 comment
This is not compatible with my device which is an LeTV X500, running Android 5.0.2. Is there any reason that it might not work? I'm in the UK, if that might be the reason.
aembleton on HN
Requests
Email or anonymous accounts
From 1 comment
Pretty nifty, but please consider allowing creation of an account with a simple email address, or even anonymously. I really don't want to have to connect my Facebook or other…
zeveb on HN
Faster logging with fewer taps
From 1 comment
I feel like its easy to make a better app if you don't try to please all use cases - just try to make the food logging in less taps. Logging in those apps takes me around 5-10…
swah on HN
Serving size control
From 1 comment
First thing I tried: logging /two/ Granny Smith apples manually. I had do the entire process twice because there is no (obvious) way to change the serving size to 2.
virtualritz on HN
Blog post or write-up of the technical approach
From 2 comments
Please do blog about the technical details, if you could post a walkthrough of setting up and training the net it would be great!
zaroth on HNIs there a writeup about how you approached this? All I see are a couple of app store links.
natch on HN
Questions
Where the training data came from
From 2 comments
The classification looks good, congrats!. Where did you get the million labeled images to train your classifier?
npalli on HNwhere did you get the annotated training data ?
misiti3780 on HN
Network details and why Caffe
From 2 comments
Any info on the network itself (number of layers, input resolution, pooling, dropout etc)?
teLeopardthy on HNwhy caffe ? why not something like h2o or tensorflow ? just curious.
sandGorgon on HN
How it handles mixed plates and hidden ingredients
From 2 comments
Great! And think about using the new data from the users to improve the system (that should probably be in your terms). How well does it work on plates with multiple foods "kinda…
swah on HNI've been thinking of this exact thing for a few years now, good to see others are interested too. I think there is HUGE potential for an app that can accurately (within a…
solaxun on HN
Alternatives named
MyFitnessPal
From 2 comments
Something like this would be a fantastic addition to MyFitnessPal. I find myself sometimes not logging food on that app because I couldn't be bothered to search for it. On the…
AJRF on HNI would pay something like $5 per day for this if it was a real person who looks at my photos, categorizes them accurately, and enters them into MyFitnessPal. I won't trust an AI…
nathan_f77 on HN
Fatsecret
From 1 comment
Your app kinda sucks now, but the detection is great. You could also do food identification as a service... I'm trying to use fatsecret/myfitnesspal for counting calories.…
swah on HN
Logameal
From 1 comment
Hey Subcomos, great work! We've been working on something pretty similar, an app called Logameal (only available in the UK). We have run into (are running into) a lot of those UI…
gchristen on HN
Tuingle
From 1 comment
Another app that is similar to this is called tuingle http://tuingle.com
johnPullman on HN
SmartPlate
From 1 comment
Theres a company in the bay area trying this approach actually! They look really interesting: https://getsmartplate.com/ Cant compete with that awesomesauce!
subcosmos on HN