We study the problem of reducing the amount of labeled training data required to train supervised classification models. We approach it by leveraging Active Learning, through sequential selection of examples which benefit the model most. Selecting examples one by one is not practical for the amount of training examples required by the modern Deep Learning models. We consider the mini-batch Active Learning setting, where several examples are selected at once. We present an approach which takes into account both informativeness of the examples for the model, as well as the diversity of the examples in a mini-batch. By using the well studied K-means clustering algorithm, this approach scales better than the previously proposed approaches, and achieves comparable or better performance.

Decision making in automated driving is highly specific to the environment and thus semantic segmentation plays a key role in recognizing the objects in the environment around the car. Pixel level classification once considered a challenging task which is now becoming mature to be productized in a car. However, semantic annotation is time consuming and quite expensive. Synthetic datasets with domain adaptation techniques have been used to alleviate the lack of large annotated datasets. In this work, we explore an alternate approach of leveraging the annotations of other tasks to improve semantic segmentation. Recently, multi-task learning became a popular paradigm in automated driving which demonstrates joint learning of multiple tasks improves overall performance of each tasks. Motivated by this, we use auxiliary tasks like depth estimation to improve the performance of semantic segmentation task. We propose adaptive task loss weighting techniques to address scale issues in multi-task loss functions which become more crucial in auxiliary tasks. We experimented on automotive datasets including SYNTHIA and KITTI and obtained 3% and 5% improvement in accuracy respectively.

Neural Processes (NPs) (Garnelo et al 2018a;b) approach regression by learning to map a context set of observed input-output pairs to a distribution over regression functions. Each function models the distribution of the output given an input, conditioned on the context. NPs have the benefit of fitting observed data efficiently with linear complexity in the number of context input-output pairs, and can learn a wide family of conditional distributions; they learn predictive distributions conditioned on context sets of arbitrary size. Nonetheless, we show that NPs suffer a fundamental drawback of underfitting, giving inaccurate predictions at the inputs of the observed data they condition on. We address this issue by incorporating attention into NPs, allowing each input location to attend to the relevant context points for the prediction. We show that this greatly improves the accuracy of predictions, results in noticeably faster training, and expands the range of functions that can be modelled.

Central Pattern Generators (CPGs) are biological neural circuits capable of producing coordinated rhythmic outputs in the absence of rhythmic input. As a result, they are responsible for most rhythmic motion in living organisms. This rhythmic control is broadly applicable to fields such as locomotive robotics and medical devices. In this paper, we explore the possibility of creating a self-sustaining CPG network for reinforcement learning that learns rhythmic motion more efficiently and across more general environments than the current multilayer perceptron (MLP) baseline models. Recent work introduces the Structured Control Net (SCN), which maintains linear and nonlinear modules for local and global control, respectively. Here, we show that time-sequence architectures such as Recurrent Neural Networks (RNNs) model CPGs effectively. Combining previous work with RNNs and SCNs, we introduce the Recurrent Control Net (RCN), which adds a linear component to the, RCNs match and exceed the performance of baseline MLPs and SCNs across all environment tasks. Our findings confirm existing intuitions for RNNs on reinforcement learning tasks, and demonstrate promise of SCN-like structures in reinforcement learning.

Creating and modeling real-world graphs is a crucial problem in various applications of engineering, biology, and social sciences; however, learning the distributions of nodes/edges and sampling from them to generate realistic graphs is still challenging. Moreover, generating a diverse set of synthetic graphs that all imitate a real network is not addressed. In this paper, the novel problem of creating diverse synthetic graphs is solved. First, we devise the deep supervised subset selection (DeepS3) algorithm; Given a ground-truth set of data points, DeepS3 selects a diverse subset of all items (i.e. data points) that best represent the items in the ground-truth set. Furthermore, we propose the deep graph representation recurrent network (GRRN) as a novel generative model that learns a probabilistic representation of a real weighted graph. Training the GRRN, we generate a large set of synthetic graphs that are likely to follow the same features and adjacency patterns as the original one. Incorporating GRRN with DeepS3, we select a diverse subset of generated graphs that best represent the behaviors of the real graph (i.e. our ground-truth). We apply our model to the novel problem of power grid synthesis, where a synthetic power network is created with the same physical/geometric properties as a real power system without revealing the real locations of the substations (nodes) and the lines (edges), since such data is confidential. Experiments on the Synthetic Power Grid Data Set show accurate synthetic networks that follow similar structural and spatial properties as the real power grid.

Richard Socher is an adjunct professor at the Stanford Computer Science Department where he obtained his PhD working on deep learning with Chris Manning and Andrew Ng. He won the best Stanford CS PhD thesis award. He is now Chief Scientist at Salesforce where he leads the company's research efforts in artificial intelligence. He previously founded MetaMind, a deep learning AI platform that analyzes, labels and makes predictions on image and text data. Richard Socher is Chief Scientist at Salesforce and an adjunct professor at the Stanford Computer Science Department.

Would you ride in an autonomous vehicle if you knew that it was subject to visual problems? How about undergo cancer treatment based on a computer interpretation of radiological images such as an x-ray, ultrasound, CT, PET, or MRI scan knowing that computer vision could easily be fooled? Computer vision has a problem–it only takes slight changes in data input to fool machine learning algorithms into "seeing" things wrong. Recent advances in computer vision are largely due to the improved pattern-recognition capabilities through deep learning, a type of machine-based learning. Machine learning is a subset of artificial intelligence where a computer is able to learn concepts from processing input data either through supervised learning where the training data is labeled, or not as in unsupervised learning or a combination without explicit programming.

Researchers have identified what may be a previously unknown human ancestor, thanks to the help of artificial intelligence. A new investigation into the genome of Asian populations has spotted the footprint of a long-ago hominid that appears to have been bred from two different species of human ancestor – Neanderthal and Denisovan. This ancient hominid, who lived tens of thousands of years ago, then bred with modern humans who arrived to Asia after the'Out of Africa' migration. It comes just months after a different team revealed the discovery of a hybrid'love child' born from a Neanderthal mother and a Denisovan father. And, the new research from the Institute of Evolutionary Biology (IBE), Centro Nacional de Análisis Genómico (CNAG-CRG) of the Centre for Genomic Regulation (CRG), and the Institute of Genomics at the University of Tartu suggests such hominid hybrids may not have been all that uncommon after all.

What were the most significant machine learning/AI advances in 2018? Let's look at all of this in some more detail. If 2017 was probably the cusp of AI fear mongering and hype (as I mentioned in last year's answer), 2018 seems to have been the year where we have started to all cool down a bit. While it is true that some figures have continued to push their message of AI fear, they have probably been too busy with other issues to make of this an important point of their agenda. At the same time, it seems like the press and others have come to peace with the idea that while self-driving cars and similar technologies are coming our way, they won't happen tomorrow.

Entrepreneurs all know that Artificial Intelligence and Machine Learning are hot topics of focus among VCs. But terms like'deep learning' and'neural networks' have been quickly absorbed into even the general public's lexicon, and today it can be difficult to tell who really knows the technology they're talking about, and who is simply pumping buzzwords. Founder Institute CEO Adeo Ressi recently sat down to discuss the future of AI and Robotics Startups and Venture Capital with a panel of renowned machine learning experts, including Jeff Schneider of Uber ATC and Carnegie Mellon University, Michael Harries of The Robotics Hub, and Birpal Sachdev, PMP, MBA of Philips Health. Founders interested in these topics should checkout our newest AI and Robotics Vertical Syndicate program in Pittsburgh led by Greg Coticchia, MBA, PC and Randy Eager https://lnkd.in/gNY82P3