41 thoughts on “Deep Learning State of the Art (2019) – MIT”

  1. Awesome video. Are there good links to videos on other developments not mentioned, e.g. in healthcare and agriculture?

  2. It could be very insightful to go back to this talk in a couple of years time, and see how these ideas developped.

  3. Lex's summary is so true:
    Stochastic gradient descent and backpropagation are still backbones of the current state of the art AI techniques. Therefore, we need some innovations to see leaps in the field.
    Thank you Lex!

  4. So, the AutoAugment is about augmentation of worst possible inputs? Like you try to do augmented ops that are hardest to recognize correctly thus force network to learn them better?

  5. I see a hand-waving review of acronyms and no step-by-step tracing of algorithms. No substance. NeXT!

  6. Thanks for the video. It will be the primary resource for our event: https://www.meetup.com/Paris-Machine-Learning-Study-Group-in-English-Meetup/events/tlzcqqyzdbhb/

  7. If this lecture is over my head and I need a little more knowledge about the fundamental concepts where should I look for that?

  8. Very happy to see the prosperity of deep learning. I hope I can excavate the biggest potential of DL in the field of computational advertising.

  9. multi complex layer systems are the future, they can be called deep complex neural networks. what I mean by that is, on each layer (to accelerate processing, inference or probable future AI applications) it is possible to train the same dataset for many different inputs, for example, it is possible to train a dataset with a video file, equipped with IMU data, with MIC input and system can learn, what type of orientation of the IMU can cause audio level breaks and what type of high-speed angle changes can cause over or under exposures. we are working on those issues with HDR capable sensors to get always-perfect image and near-perfect audio with mutiple level microphone inputs. similar techniques can be applied to autonomous driving, medical robots or self driving rovers on moon or mars.

  10. adanet is very interesting! that is a very good material on progressive learning for data scarse situation! also multiple classification in rounds… and one on confusion for fine grain classifications without augmentation!

  11. I was absolutely shocked by how brilliant these approaches to deep learning where. I'm absolutely excited to see what we can come up with next

  12. BLOW MIND!!!!. I have a question. To simulate the human brain you also may need to simulated the human body because many of those connection I believe comes from the organs that the brain is sustaining to keep it alive. Plus in the human body depending of the blood flow, the toxin of the blood, the well of the organs, the air that the lung use, almost everything on the organ affect the function of the brain work. Plus, with the 100 billion neuron with more than 1 trillion connection in the brain. SO, do you think that after all that research of the human connection it also has to be done a research of the human body to simulate inside a computer?.

  13. Great review on recent advance in deep learning! Would be great to see a similar review of current (immediate) challenges e.g. limited numerical extrapolation abilities, multi-task learning…etc.

  14. This is attempting computational expression! It inhibits natural inherent algorithmic processing. Human is an emotional cognitive being. Whether it be locution or calculation, induction/deduction is based on emotional cognitive aptitude, therewith conflict with the attempt of mimicking sterile computation and natural anatomization. I have watched from afar and Chomsky is aware who I am… ( ^V^ ) If human continues to use technology without the comprehension that tech creates an ersatz structure, than human will fall deeper and deeper into a form of psychosis. Love to ALL

  15. Would you please make a video detailed lecture on computer vision and algorithm for real time tracking applications?

  16. This is fantastic. I am trying to create a storytelling system using LSTM s and a corpus of self-written works. Thank you for this Mr. Fridman.

  17. Thanks for the lecture. Also, well edited (erasing pauses). Funny to see you transform to Men in Blue compared to when you started the lecture two years ago. Looking good

  18. Thanks for the lecture. Also, well edited (erasing pauses). Funny to see you transform to Men in Blue compared to when you started the lecture two years ago. Looking good

  19. I'm excited about recent developments in NLP, deep RL, speeding up training/inference, big GANs, powerful DL frameworks, and real-world application of DL in driving 370k+ Tesla HW2 cars! What else would you like to see covered?

Leave a Reply

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