How we teach computers to understand pictures | Fei Fei Li



When a very young child looks at a picture, she can identify simple elements: “cat,” “book,” “chair.” Now, computers are getting smart enough to do that too. What’s next? In a thrilling talk, computer vision expert Fei-Fei Li describes the state of the art — including the database of 15 million photos her team built to “teach” a computer to understand pictures — and the key insights yet to come.

TEDTalks is a daily video podcast of the best talks and performances from the TED Conference, where the world’s leading thinkers and doers give the talk of their lives in 18 minutes (or less). Look for talks on Technology, Entertainment and Design — plus science, business, global issues, the arts and much more.
Find closed captions and translated subtitles in many languages at http://www.ted.com/translate

Follow TED news on Twitter: http://www.twitter.com/tednews
Like TED on Facebook: https://www.facebook.com/TED

Subscribe to our channel: http://www.youtube.com/user/TEDtalksDirector

source

Fahad Hameed

Fahad Hashmi is one of the known Software Engineer and blogger likes to blog about design resources. He is passionate about collecting the awe-inspiring design tools, to help designers.He blogs only for Designers & Photographers.

47 thoughts on “How we teach computers to understand pictures | Fei Fei Li

  • August 30, 2017 at 10:25 pm
    Permalink

    the dress that she is wearing is also made of pictures!. amazing effort she took to make image recognition by computers possible.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    a child only needs to see ONE cat to pretty much immediately identify all the other cats it will see in its lifetime. the training given to the NN seems very different to this: feed it umpteen "different" cat images.

    the child, only needs a few images of the same cat .. PLUS .. non-visual data (feel, sounds,real-time interaction). Not to mention that a human has two eyes separated by a distance yielding a 3D-view .. these are more complex images that have not been fed to the NN.

    so, the result of the NN we have seen is indeed impressive .. but it seems the inputs to the NN are VERY different to the human experience.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    Well… this is flawed in that a human child doesn't just look at individual photos, but rather a continuous stream of images from their eyes as they look at something from different angles. There is no need for a child's mind to "stitch" random fragmented images together, because they can directly see how the shape changes as their view changes.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    Thank you Dr. Fei, well said , peace and love, Doug :). PS The dress you were wearing Is incredible !!

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    Eliza might say we are anthropomorphiseing her words. Maybe a boy at a table with a cake is less important than a boy with an elevated temperature. Why not use all available data, hyperspectral imaging. I typed Eliza into a 1MHz Z80 computer in 1977. Alexa could take a few tips from her.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    Technology is becoming more like the god of today society how sad. I can't believe she is talking about computers surpassing the human brain capability, how sick some people are in this world.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    Technology is amazing, but also overrated when people talk of all the amazing benefits that technology brings to our lives and how it's going to solve the world's problems. The worlds most complex problems are predominantly human problems not technology problems.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    Only thing kept me listening is that Big brother not only watches but also alert or even save a life

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    How to participate in such crowd-sourcing projects? can anyone suggest a portal or social media page which gives notifications etc.?

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    stupid white girls and asian girls work hard and don't marry and die leaving nothing behind them. black girls and latina are far smarter than whites and asian girls.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    I am a computer science under-grad student. How can I start with computer vision?

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    Google is on the wrong track with deep learning. Numenta is on the right track with its HTM model. Time will tell who is correct.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    she's so cute drawing inspiration for this stuff from observing her children grow

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    also, if we take advancements on robotics into account then birth of Skynet will become a reality in the near future.

    Reply
  • August 30, 2017 at 10:25 pm
    Permalink

    Her voice is so high and piercing that makes me uncomfortable ! wanna pee~ ! (no offense)

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *