Life is a game, take it seriously

One dollar classifier? NEIL, the never ending image learner

In Computer Vision, Machine Learning on November 27, 2013 at 5:18 pm

by Li Yang Ku (Gooly)

NEIL never ending image learner

I had the chance to chat with Abhinav Gupta, a research professor at CMU, in person when he visited UMass Amherst about a month ago. Abhinav presented NEIL, the never ending image learner in his talk at Amherst. To give a short intro, the following is from Abhinav

“NEIL (Never Ending Image Learner) is a computer program that runs 24 hours per day and 7 days per week to automatically extract visual knowledge from Internet data. It is an effort to build the world’s largest visual knowledge base with minimum human labeling effort – one that would be useful to many computer vision and AI efforts.” 

NEIL never ending image learner clusters

One of the characteristic that distinguishes NEIL from other object recognition algorithms that are trained and tested on large web image data set such as the ImageNet or LFW is that NEIL is trying to recognize images that are in a set that has unlimited data and unlimited category. At first glance this might look like a problem too hard to solve. But NEIL approaches this problem in a smart way. Instead of trying to label images one by one on the internet, NEIL start from labeling just the easy ones. Since given a keyword the number of images returned are so large using Google Image Search, NEIL simply picks the ones it feels most certain, which are the ones that share the most common HOG like features. This step also helps refining the query result. Say we searched for cars on Google Image, it is very likely that out of every 100 images there is one image that has nothing to do with cars (very likely some sexy photo of girls with file name girl_love_cars.jpg ). These outliers won’t share the same visual features as the other car clusters and will not be labeled. By doing so NEIL can gradually build  up a very large labeled data set from one word to another.

girl_love_car

NEIL also learns the relationships between images and is connected with NELL, never ending language learning. More details should be released in future papers. During the talk Abhinav said he plan to set up a system where you can submit the category you wanna train on and with just $1, NEIL will give you a set of HOG classifiers in that category in 1 day.

NEIL relationship

  1. This seems like a cool project but I wonder how NEIL decides where to draw a decision boundary between two categories? Initially, for easy exemplars it’s easy, but as more complex relations must be inferred (e.g., objects from less common viewpoints), it becomes much less clear how to cluster them. But perhaps that’s not the goal? Perhaps they only want to extract the (easy) prototypical images for each category?

    • Well, I can’t speak for the NEIL team, but if there is an image taken from a weird viewpoint even human will have a hard time recognizing it. I think the assumption is that images of an object taken from any reasonable angle exist or will exist on the internet with more than one copy. So even if it doesn’t appear on Google image search, it can be recovered by finding the most similar image taken from similar viewpoint. As long as no false positive is included, it should be able to find more and more images taken from different viewpoint gradually. Think of “build Rome in a day” https://computervisionblog.wordpress.com/2011/04/25/surf-on-images-feature-point-matching/

  2. this is nice blog and very impressive please use this thanks
    for this blog

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: