Yann LeCun

Yann LeCun is a French-American data scientist born on 8 July 1960. He is also known as the father of convolutional neural networks. He is the Silver professor of the Courant institute of Mathematical Sciences at New York university. He is also the vice president of the chief AI scientist at Facebook.

He is the founding father of convolutional nets and using those he has done man projects : 1. Optical character recognition; 2.Computer vision; 3.Mobile robotics; etc. He co-developed the Lush programming language along with Leon Bottou. LeCun along with Geoffery Hinton and Yoshua Bengio are reffered as “Godfathers of AI “ and “Godfathers of deep learning”.

He also won the ACM A.M. Turing award for his work in deep learning in 2018. Optical character recognition (one of Yann’s great works) helps computers read text. This is a very helpful feature as many people at the time of his invention were using scanners which cause many problems like the scanner screen was dirty and more. With the help of this technology anyone can use scan text with ease and moreover it has also been trained to recognize the handwriting of people and it is also able to scan text written in bad handwriting or blur. This saved a lot of time and money. Computer vision (another great work of LeCun) helps to scan images and videos. This has helped others to create other technologies using this tech like, for instance, cst scanner which detects images of different animals and tells which is the image of a cat. This great technology is available only because of Yann’s computer vision technology.

We should be so grateful of him because it is because of him that we are able to perform different tasks with ease. He also proposed the method of backpropagation in neural networks to help adjust the weights and get the minimum error rate. This has been really helpful in building all the good neural networks that are now present in this world. Backpropagation means that we work in the opposite manner.

Instead of giving the image of a cat or dog, for instance, and asking it to tell whether it is a dog or cat we show just show it the image and tell the network that it is a cat or dog. Likewise we give many images like this and the network adjusts the weights accordingly and after running a few images in opposite order we have our network ready to be used with almost correct weights. He also said that we shouldn’t run images of a very particular type rather in more general terms so that the AI is well trained to tell the correct answers even in the images it hadn’t seen before.He has also contributed in face recognition development. He also made a network that could detect objects and tell the name of the object.

Yann has also worked on an automatic signature verification system based on Siamese network that would detect the signature using distance between objects. He also applied this technology to Chinese language. Though he later left the handwriting model because he left that it was an ineffective way of communication , he learned how to train a large and complex system at the world level without requiring manual segmentation. Yann has also won the IEEE Neural Network Pioneer Award in 2014 PAMI Distinguished Researcher Award in 2015. In 2017 he was invited to lecture at King Abdullah University of Science and Technology in Saudi Arabia but he declined it as he thought that they might consider him a terrorist in the country in view of his atheism. In 2019 h e won the Turing Award along with Yoshua Bengio and Geoffrey Hinton.

Yann has made wonderful inventions in the field of machine learning and they have been helping us since a long time and we hope that he will continue to make more inventions.