MIT's introductory course on deep learning methods with applications to computer vision, natural language processing, biology, and more! Students will gain foundational knowledge of deep learning algorithms and get practical experience in building neural networks in TensorFlow. Course concludes with a project proposal competition with feedback from staff and panel of industry sponsors. Prerequisites assume calculus (i.e. taking derivatives) and linear algebra (i.e. matrix multiplication), we'll try to explain everything else along the way! Experience in Python is helpful but not necessary. Listeners are welcome!
1:00pm - 4:00pm EST Everyday
1:00pm-1:45pm: Lecture Part 1
1:45pm-2:30pm: Lecture Part 2
2:30pm-4:00pm: Software Labs
Classes will take place online virtually via MIT Canvas and publically open-sourced every week, starting March 9th, 2022.
[Code]
[Code]
We are expecting very elementary knowledge of linear algebra and calculus. How to multiply matrices, take derivatives and apply the chain rule. Familiarity in Python is a big plus as well. The course will be beginner friendly since we have many registered students from outside of computer science.
If you would like to receive course related updates and lecture materials please subscribe to our YouTube channel and sign up for our mailing list.
All course materials available online for free but are copyrighted and licensed under the MIT license. If you are an instructor and would like to use any materials from this course (slides, labs, code), you must add the following reference to each slide:
© Alexander Amini and Ava Soleimany
MIT 6.S191: Introduction to Deep Learning
IntroToDeepLearning.com
All course materials are copyrighted and licensed under the MIT license. If you are an instructor and would like to use any materials from this course (slides, labs, code), you must add the following reference to each slide:
© Alexander Amini and Ava Soleimany
MIT 6.S191: Introduction to Deep Learning
IntroToDeepLearning.com
If you are an MIT student, postdoc, faculty, or affiliate and would like to become involved with this course please email introtodeeplearning-staff@mit.edu. We are always accepting new applications to join the course staff.
This class would not be possible without our amazing sponsors and has been sponsored by Google, IBM, NVIDIA, Ernst and Young, LambdaLabs, Tencent AI, Microsoft, Amazon, and Onepanel. If you are interested in becoming involved in this course as a sponsor please contact us at introtodeeplearning-staff@mit.edu.
Copyright © MIT 6.S191. banner image; page template
Mass production is a major challenge for automakers set on delivering safe automated driving. For example, they must consider impaired vision caused by weather, lighting, and other adverse road conditions, including slopes and curvatures from potholes or hills. Innoviz CEO Omer Keilaf will offer an introduction to automotive requirements that carmakers ask for in order to meet their targets. Amir Day, Innoviz director of computer vision, will then match these requirements with accompanying perception software challenges and share experiences from how Innoviz solves them.
Omer Keilaf has spent the greater part of the past two decades driving cutting edge technologies from inception to commercialization. As Co-Founder and CEO of Innoviz Technologies, Omer is working tirelessly to bring safe AVs to consumer markets by delivering high-performance, automotive-grade LiDAR and perception software to automakers at mass market prices. Last year, Keliaf took his company public through a SPAC merger with Collective Growth Corporation, raising over $370 million. Prior to founding Innoviz, Omer served as an officer in the elite IDF intelligence Unit 81, and held senior leadership roles at various companies. Omer holds a BSc and MSc in Electrical Engineering and an MBA, all from Tel Aviv University.
Amir Day is the Director for Computer Vision at Innoviz Technologies. With more than a decade of experience executing complex, multi-disciplinary technology projects from conception to production, Amir's team developed Innoviz's perception software (InnovizAPP) from scratch, focusing on deep learning and software infrastructure as a means to interpret Innoviz's rich point cloud data. Before working at Innoviz, Amir served in the IDF's Military Intelligence Corps, leading a team of 100+ engineers to develop new, novel systems solutions. Today, Amir focuses on 3D Computer Vision, Deep Learning and Data Pipeline for autonomous vehicles perception.
Rev.com is one of the world's largest transcription companies by volume, processing over 2500 hours of audio and video every day. Rev has turned transcription into a "virtuous cycle" between human transcribers and automatic speech recognition (ASR) - automatically-generated first drafts make humans more efficient and human-edited transcripts are used to train better and better ASR models. In this talk, we will discuss the state-of-the-art in deep neural network models for speech recognition and how Rev has built the world's best English ASR engine. We will cover several different model architectures for ASR, exploring the impact of modeling choices on accuracy, speed, and the ease of implementing additional features like word timings and customization. We will end with an overview of Rev's current research efforts and future plans.
Miguel Jetté is the Head of AI Research and Development at Rev AI. Miguel has over 20 years of experience in speech recognition and machine learning, and holds degrees in Mathematics and Computer Science from McGill University. Miguel is passionate about leveraging mathematics, computer science, statistics, and their intersection to solve important problems that improve the quality of human life.
Jennifer Drexler is a Senior Speech Scientist at Rev, where she works on problems in speech recognition and language processing. She completed her PhD in Electrical Engineering and Computer Science at MIT, where she worked on enabling speech technologies in low-resource settings, and has worked on machine translation in the Human Language Technology Group at MIT’s Lincoln Labs.
coming soon!
Anima Anandkumar is a Bren professor in Computing and Mathematical Sciences at Caltech, and a director of machine learning research at NVIDIA. At Caltech, she is the co-director of the Design, Optimization, and Learning initative and co-leads the AI4science initiative. At NVIDIA, she is leading a research group that develops next-generation AI algorithms. Amongst her many scientific contributions, she spearheaded the development of tensor algorithms, which are are central to achieving massive parallelism in large-scale AI applications. She has been recognized by several awards including an Alfred. P. Sloan Fellowship, NSF Career Award, and faculty fellowships from Microsoft, Google, and Adobe. Anima received her B.Tech in Electrical Engineering from IIT Madras in 2004 and her PhD from Cornell University in 2009, and completed postdoctoral research at MIT.
Deep learning models are bad at signalling failure: They tend to make predictions with high confidence, and this is problematic in real-world applications such as healthcare, self-driving cars, and natural language systems, where there are considerable safety implications, or where there are discrepancies between the training data and data that the model makes predictions on. There is a pressing need both for understanding when models should not make predictions and improving model robustness to natural changes in the data. In this talk, I'll give a very abridged version of my NeurIPS tutorial on uncertainty and robustness in deep learning and then introduce some more recent work developed to address these challenges.
Jasper Snoek is currently a staff research scientist at Google Brain. Recently his research has focused on methods for improving uncertainty and robustness of deep learning methods. His interests span a variety of topics at the intersection of Bayesian methods and deep learning. He completed his PhD in machine learning at the University of Toronto. He subsequently held postdoctoral fellowships at the University of Toronto, under Geoffrey Hinton and Ruslan Salakhutdinov, and at the Harvard Center for Research on Computation and Society, under Ryan Adams. Jasper co-founded the machine learning startup Whetlab, which was acquired by Twitter. He has served as an Area Chair for NeurIPS, ICML and ICLR, and organized a variety of workshops at ICML and NeurIPS.
