285

03:29

Two Minute Papers: Training Deep Neural Networks With Dropout | Two Minute Papers #62

In this episode, we discuss the bane of many machine learning algorithms – overfitting. It is also explained why it is an undesirable way to learn and how to combat it via dropout. _____________________ The paper “Dropout: A Simple Way to Prevent Neural Networks from Overtting” is available here: https://www.cs.toronto.edu/~hinton/absps/JMLRdropout.pdf Andrej Karpathy’s autoencoder is available […]

0

304

04:34

Two Minute Papers: Overfitting and Regularization For Deep Learning | Two Minute Papers #56

In this episode, we discuss the bane of many machine learning algorithms – overfitting. It is also explained why it is an undesirable way to learn and how to combat it via L1 and L2 regularization. _____________________________ The paper “Regression Shrinkage and Selection via the Lasso” is available here: http://statweb.stanford.edu/~tibs/lasso/lasso.pdf Andrej Karpathy’s excellent lecture notes […]

0

314

03:25

Two Minute Papers: Breaking Deep Learning Systems With Adversarial Examples | Two Minute Papers #43

Artificial neural networks are computer programs that try to approximate what the human brain does to solve problems like recognizing objects in images. In this piece of work, the authors analyze the properties of these neural networks and try to unveil what exactly makes them think that a paper towel is a paper towel, and, […]

0

0

03:10

CoRL 2020, Spotlight Talk 184: MuGNet: Multi-Resolution Graph Neural Network for Segmenting Large…

“**MuGNet: Multi-Resolution Graph Neural Network for Segmenting Large-Scale Pointclouds** Liuyue Xie (Carnegie Mellon University)*; Tomotake Furuhata (Carnegie Mellon University); Kenji Shimada (Carnegie Mellon University) Publication: http://corlconf.github.io/paper_184/ **Abstract** In this paper, we propose a multi-resolution deep-learning architecture to segment dense large-scale pointclouds semantically. Dense pointcloud data require a computationally expensive feature encoding process before semantic segmentation. […]

0

0

05:05

CoRL 2020, Spotlight Talk 381: Learning Object-conditioned Exploration using Distributed Soft Act…

“**Learning Object-conditioned Exploration using Distributed Soft Actor Critic** Ayzaan Wahid (Google)*; Austin C Stone (Google); Kevin Chen (Stanford); Brian Ichter (Google Brain); Alexander Toshev (Google) Publication: http://corlconf.github.io/paper_381/ **Abstract** Object navigation is defined as navigating to an object of a given label in a complex, unexplored environment. In its general form, this problem poses several challenges […]

0

0

05:04

CoRL 2020, Spotlight Talk 45: Positive-Unlabeled Reward Learning

“**Positive-Unlabeled Reward Learning** Danfei Xu (Stanford University)*; Misha Denil (DeepMind) Publication: http://corlconf.github.io/paper_45/ **Abstract** Learning reward functions from data is a promising path towards achieving scalable Reinforcement Learning (RL) for robotics. However, a major challenge in training agents from learned reward models is that the agent can learn to exploit errors in the reward model to […]

0

1

05:05

CoRL 2020, Spotlight Talk 459: Sampling-based Reachability Analysis: A Random Set Theory Approach…

“**Sampling-based Reachability Analysis: A Random Set Theory Approach with Adversarial Sampling** Thomas Lew (Stanford University)*; Marco Pavone (Stanford University) Publication: http://corlconf.github.io/paper_459/ **Abstract** Reachability analysis is at the core of many applications, from neural network verification, to safe trajectory planning of uncertain systems. However, this problem is notoriously challenging, and current approaches tend to be either […]

0

524

02:41

Two Minute Papers: Recurrent Neural Network Writes Sentences About Images | Two Minute Papers #23

This technique is a combination of two powerful machine learning algorithms: – convolutional neural networks are excellent at image classification, i.e., finding out what is seen on an input image, – recurrent neural networks that are capable of processing a sequence of inputs and outputs, therefore it can create sentences of what is seen on […]

0

613

02:54

Two Minute Papers: Image Colorization With Deep Learning and Classification | Two Minute Papers #71

The paper “Let there be Color!: Joint End-to-end Learning of Global and Local Image Priors for Automatic Image Colorization with Simultaneous Classification” and its implementation are available here: http://hi.cs.waseda.ac.jp/~iizuka/projects/colorization/en/ https://github.com/satoshiiizuka/siggraph2016_colorization The video classification paper by Karpathy et al.: http://cs.stanford.edu/people/karpathy/deepvideo/ Recommended for you: Artistic Style Transfer For Videos – https://www.youtube.com/watch?v=Uxax5EKg0zA Deep Learning related Two Minute Papers […]

0

2

05:20

CoRL 2020, Spotlight Talk 499: MATS: An Interpretable Trajectory Forecasting Representation for P…

“**MATS: An Interpretable Trajectory Forecasting Representation for Planning and Control** Boris Ivanovic (Stanford University)*; Amine Elhafsi (Stanford University); Guy Rosman (Toyota Research Institute); Adrien Gaidon (Toyota Research Institute); Marco Pavone (Stanford University) Publication: http://corlconf.github.io/paper_499/ **Abstract** Reasoning about human motion is a core component of modern human-robot interactive systems. In particular, one of the main uses […]

0

0

05:06

CoRL 2020, Spotlight Talk 526: Robust Policies via Mid-Level Visual Representations: An Experimen…

**Robust Policies via Mid-Level Visual Representations: An Experimental Study in Manipulation and Navigation** Bryan Chen (UC Berkeley)*; Alexander Sax (UC Berkeley); Francis Lewis (Stanford); Iro Armeni (Stanford University); Silvio Savarese (Stanford University); Amir Zamir (Swiss Federal Institute of Technology (EPFL)); Jitendra Malik (University of California at Berkeley); Lerrel Pinto (NYU/Berkeley) Publication: http://corlconf.github.io/paper_526/ **Abstract** Vision-based robotics […]

0

1

05:05

CoRL 2020, Spotlight Talk 408: Safe Optimal Control Using Stochastic Barrier Functions and Deep F…

“**Safe Optimal Control Using Stochastic Barrier Functions and Deep Forward-Backward SDEs** Marcus Pereira (Georgia Institute Technology)*; Ziyi Wang (Georgia Institute of Technology); Ioannis Exarchos (Stanford University); Evangelos Theodorou (Georgia Institute of Technology) Publication: http://corlconf.github.io/paper_408/ **Abstract** This paper introduces a new formulation for stochastic optimal control and stochastic dynamic optimization that ensures safety with respect to […]

0

2.11K

03:42

Two Minute Papers: This is How You Hack A Neural Network

The paper “Adversarial Reprogramming of Neural Networks” is available here: https://arxiv.org/abs/1806.11146 Andrej Karpathy’s image classifier: https://cs.stanford.edu/people/karpathy/convnetjs/demo/cifar10.html Pick up cool perks on our Patreon page: https://www.patreon.com/TwoMinutePapers We would like to thank our generous Patreon supporters who make Two Minute Papers possible: 313V, Andrew Melnychuk, Angelos Evripiotis, Brian Gilman, Christian Ahlin, Christoph Jadanowski, Dennis Abts, Emmanuel, Eric […]

0

2.14K

04:32

Two Minute Papers: This Neural Network Optimizes Itself | Two Minute Papers #212

The paper “Hierarchical Representations for Efficient Architecture Search” is available here: https://arxiv.org/pdf/1711.00436.pdf Genetic algorithm (+ Mona Lisa problem) implementation: 1. https://users.cg.tuwien.ac.at/zsolnai/gfx/mona_lisa_parallel_genetic_algorithm/ 2. https://users.cg.tuwien.ac.at/zsolnai/gfx/knapsack_genetic/ Andrej Karpathy’s online demo: http://cs.stanford.edu/people/karpathy/convnetjs/demo/classify2d.html Overfitting and Regularization For Deep Learning – https://www.youtube.com/watch?v=6aF9sJrzxaM Training Deep Neural Networks With Dropout – https://www.youtube.com/watch?v=LhhEv1dMpKE How Do Genetic Algorithms Work? – https://www.youtube.com/watch?v=ziMHaGQJuSI We would like to […]

0

3

05:02

CoRL 2020, Spotlight Talk 66: Reinforcement Learning with Videos: Combining Offline Observations …

“**Reinforcement Learning with Videos: Combining Offline Observations with Interaction** Karl Schmeckpeper (University of Pennsylvania)*; Oleh Rybkin (University of Pennsylvania); Kostas Daniilidis (University of Pennsylvania); Sergey Levine (UC Berkeley); Chelsea Finn (Stanford) Publication: http://corlconf.github.io/paper_66/ **Abstract** Reinforcement learning is a powerful framework for robots to acquire skills from experience, but often requires a substantial amount of online […]

0

6

05:05

CoRL 2020, Spotlight Talk 128: Learning Latent Representations to Influence Multi-Agent Interaction

“**Learning Latent Representations to Influence Multi-Agent Interaction** Annie Xie (Stanford University)*; Dylan Losey (Stanford University); Ryan Tolsma (Stanford University); Chelsea Finn (Stanford); Dorsa Sadigh (Stanford) Publication: http://corlconf.github.io/paper_128/ **Abstract** Seamlessly interacting with humans or robots is hard because these agents are non-stationary. They update their policy in response to the ego agent’s behavior, and the ego […]

0

1.38K

04:02

Two Minute Papers: 10 Even Cooler Deep Learning Applications | Two Minute Papers #59

For the third time, we present another round of incredible deep learning applications! ___________________ 1. Geolocation – http://arxiv.org/abs/1602.05314 2. Super-resolution – http://arxiv.org/pdf/1511.04491v1.pdf 3. Neural Network visualizer – http://experiments.mostafa.io/public/ffbpann/ 4. Recurrent neural network for sentence completion: http://www.cs.toronto.edu/~ilya/fourth.cgi 5. Human-in-the-loop and Doctor-in-the-loop: http://link.springer.com/article/10.1007/s40708-016-0036-4 6. Emoji suggestions for images – https://emojini.curalate.com/ 7. MNIST handwritten numbers in HD – […]

0

9

45:06

CoRL 2020 Keynote 2 – Dorsa Sadigh

Walking the Boundary of Learning and Interaction Dorsa Sadigh (Stanford) More information: https://www.robot-learning.org/program/keynotes YouTube Source for this Robot AI Video

0

5

04:33

CoRL 2020, Spotlight Talk 189: TNT: Target-driveN Trajectory Prediction

“**TNT: Target-driveN Trajectory Prediction** Hang Zhao (Waymo)*; Jiyang Gao (Waymo); Tian Lan (Waymo); Chen Sun (Google); Ben Sapp (Waymo); Balakrishnan Varadarajan (Google Research); Yue Shen (Waymo, LLC); Yi Shen (Waymo); Yuning Chai (Waymo); Cordelia Schmid (Google); Congcong Li (Waymo); Dragomir Anguelov (Waymo) Publication: http://corlconf.github.io/paper_189/ **Abstract** Predicting the future behavior of moving agents is essential for […]

0

4.01K

04:57

Adversarial Attacks on Neural Networks – Bug or Feature?

❤️ Support us on Patreon: https://www.patreon.com/TwoMinutePapers 📝 The paper “Adversarial Examples Are Not Bugs, They Are Features” is available here: http://gradientscience.org/adv/ The Distill discussion article is available here: https://distill.pub/2019/advex-bugs-discussion/ If you wish to play with some of these Distill articles, look here: – https://distill.pub/2017/feature-visualization/ – https://distill.pub/2018/building-blocks/ Andrej Karpathy’s image classifier – you can run this […]

0

2.67K

03:50

What is an Autoencoder? | Two Minute Papers #86

Autoencoders are neural networks that are capable of creating sparse representations of the input data and can therefore be used for image compression. There are denoising autoencoders that after learning these sparse representations, can be presented with noisy images. What is even better is a variant that is called the variational autoencoder that not only […]

0

0

T12 06 Panel: Challenges and opportunities of causality

Moderator: Eric Horvitz, Chief Scientific Officer, Microsoft Speakers: Yoshua Bengio, Scientific Director / Full Professor, Université de Montréal Susan Athey, Professor, Stanford University Judea Pearl, Professor, UCLA What is causal machine learning? Is it the same as causality research? What are the recent advances and future opportunities? This panel is a unique session where you […]

0

0

These Natural Images Fool Neural Networks (And Maybe You Too)

❤️ Check out Weights & Biases here and sign up for a free demo: https://www.wandb.com/papers Their blog post on training a neural network is available here: https://www.wandb.com/articles/mnist 📝 The paper “Natural Adversarial Examples” and its dataset are available here: https://arxiv.org/abs/1907.07174 https://github.com/hendrycks/natural-adv-examples Andrej Karpathy’s image classifier: https://cs.stanford.edu/people/karpathy/convnetjs/demo/cifar10.html You can also join us here to get early […]

0

281

01:16:40

Lecture 20 | Machine Learning (Stanford)

Lecture by Professor Andrew Ng for Machine Learning (CS 229) in the Stanford Computer Science department. Professor Ng discusses POMDPs, policy search, and Pegasus in the context of reinforcement learning. This course provides a broad introduction to machine learning and statistical pattern recognition. Topics include supervised learning, unsupervised learning, learning theory, reinforcement learning and adaptive […]

0