We demonstrate a conditional autoregressive pipeline for efficient music recomposition, based on methods presented in van den Oord et al.(2017). Recomposition (Casal & Casey, 2010) focuses on reworking existing musical pieces, adhering to structure at a high level while also re-imagining other aspects of the work. This can involve reuse of pre-existing themes or parts of the original piece, while also requiring the flexibility to generate new content at different levels of granularity. Applying the aforementioned modeling pipeline to recomposition, we show diverse and structured generation conditioned on chord sequence annotations.

Deep Neural Networks (DNNs) have become a powerful tool for a wide range of problems. Yet recent work has shown an increasing variety of adversarial samples that can fool them. Most existing detection mechanisms impose significant costs, either by using additional classifiers to spot adversarial samples, or by requiring the DNN to be restructured. In this paper, we introduce a novel defence. We train our DNN so that, as long as it is working as intended on the kind of inputs we expect, its behavior is constrained, in that a set of behaviors are taboo. If it is exposed to adversarial samples, they will often cause a taboo behavior, which we can detect. As an analogy, we can imagine that we are teaching our robot good manners; if it's ever rude, we know it's come under some bad influence. This defence mechanism is very simple and, although it involves a modest increase in training, has almost zero computation overhead at runtime -- making it particularly suitable for use in embedded systems. Taboos can be both subtle and diverse. Just as humans' choice of language can convey a lot of information about location, affiliation, class and much else that can be opaque to outsiders but that enables members of the same group to recognise each other, so also taboo choice can encode and hide information. We can use this to make adversarial attacks much harder. It is a well-established design principle that the security of a system should not depend on the obscurity of its design, but of some variable (the key) which can differ between implementations and be changed as necessary. We explain how taboos can be used to equip a classifier with just such a key, and to tune the keying mechanism to adversaries of various capabilities. We evaluate the performance of a prototype against a wide range of attacks and show how our simple defense can work well in practice.

The past few years have witnessed the fast development of different regularization methods for deep learning models such as fully-connected deep neural networks (DNNs) and Convolutional Neural Networks (CNNs). Most of previous methods mainly consider to drop features from input data and hidden layers, such as Dropout, Cutout and DropBlocks. DropConnect select to drop connections between fully-connected layers. By randomly discard some features or connections, the above mentioned methods control the overfitting problem and improve the performance of neural networks. In this paper, we proposed two novel regularization methods, namely DropFilter and DropFilter-PLUS, for the learning of CNNs. Different from the previous methods, DropFilter and DropFilter-PLUS selects to modify the convolution filters. For DropFilter-PLUS, we find a suitable way to accelerate the learning process based on theoretical analysis. Experimental results on MNIST show that using DropFilter and DropFilter-PLUS may improve performance on image classification tasks.

We present PredRNN++, an improved recurrent network for video predictive learning. In pursuit of a greater spatiotemporal modeling capability, our approach increases the transition depth between adjacent states by leveraging a novel recurrent unit, which is named Causal LSTM for re-organizing the spatial and temporal memories in a cascaded mechanism. However, there is still a dilemma in video predictive learning: increasingly deep-in-time models have been designed for capturing complex variations, while introducing more difficulties in the gradient back-propagation. To alleviate this undesirable effect, we propose a Gradient Highway architecture, which provides alternative shorter routes for gradient flows from outputs back to long-range inputs. This architecture works seamlessly with causal LSTMs, enabling PredRNN++ to capture short-term and long-term dependencies adaptively. We assess our model on both synthetic and real video datasets, showing its ability to ease the vanishing gradient problem and yield state-of-the-art prediction results even in a difficult objects occlusion scenario.

With the increasing number of base stations (BSs) and network densification in 5G, interference management using link scheduling and power control are vital for better utilization of radio resources. However, the complexity of solving link scheduling and the power control problem grows exponentially with the number of BS. Due to high computation time, previous methods are useful for research purposes but impractical for real time usage. In this paper we propose to use deep neural networks (DNNs) to approximate optimal link scheduling and power control for the case with multiple small cells. A deep Q-network (DQN) estimates a suitable schedule, then a DNN allocates power for the corresponding schedule. Simulation results show that the proposed method achieves over five orders of magnitude speed-up with less than nine percent performance loss, making real time usage practical.

Humans make accurate decisions by interpreting complex data from multiple sources. Medical diagnostics, in particular, often hinge on human interpretation of multi-modal information. In order for artificial intelligence to make progress in automated, objective, and accurate diagnosis and prognosis, methods to fuse information from multiple medical imaging modalities are required. However, combining information from multiple data sources has several challenges, as current deep learning architectures lack the ability to extract useful representations from multimodal information, and often simple concatenation is used to fuse such information. In this work, we propose Multimodal DenseNet, a novel architecture for fusing multimodal data. Instead of focusing on concatenation or early and late fusion, our proposed architectures fuses information over several layers and gives the model flexibility in how it combines information from multiple sources. We apply this architecture to the challenge of polyp characterization and landmark identification in endoscopy. Features from white light images are fused with features from narrow band imaging or depth maps. This study demonstrates that Multimodal DenseNet outperforms monomodal classification as well as other multimodal fusion techniques by a significant margin on two different datasets.

With the recent surge of interests in deep neural networks, more real-world applications start to adopt it in practice. However, deep neural networks are known to have limited control over its prediction under unseen images. Such weakness can potentially threaten society and cause annoying consequences in real-world scenarios. In order to resolve such issue, a popular task called out-of-distribution detection was proposed, which aims at separating out-of-distribution images from in-distribution images. In this paper, we propose a perturbed prior network architecture, which can efficiently separate model-level uncertainty from data-level uncertainty via prior entropy. To further enhance the robustness of proposed entropy-based uncertainty measure, we propose a concentration perturbation algorithm, which adaptively adds noise to concentration parameters so that the in- and out-of-distribution images are better separable. Our method can directly rely on the pre-trained deep neural network without re-training it, and also requires no knowledge about the network architecture and out-of-distribution examples. Such simplicity makes our method more suitable for real-world AI applications. Through comprehensive experiments, our methods demonstrate its superiority by achieving state-of-the-art results on many datasets.

Today we're excited to announce the public launch of MissingLink.ai to help data scientists and engineers streamline and automate the entire deep learning cycle. With this launch, we hope to eliminate a lot of the grunt work associated with machine learning and to accelerate the time it takes to train and deliver effective models. Work on MissingLink began in 2016, when my colleagues Shay Erlichmen, Rahav Lussato, and I set out to solve a problem we experienced as software engineers. While working on deep learning projects at our previous company, we realized we were spending too much time managing the sheer volume of data we were collecting and analyzing, and too little time learning from it. We also realized we weren't alone.

Thomson Reuters has a series, AI experts, where they interview thought leaders from different areas - including technology executives, researchers, robotics experts and policymakers - on what we might expect as we move towards AI. As part of that series I recently spoke to Paul Thies of Thomson Reuters, and here are the excerpts from the interview: Anticipating the next move in data science Thomson Reuters: For timely information concerning developments in data science, data mining and business analytics, KDnuggets is widely regarded as a leading outlet in the field. Created in 1993 by founder, editor and president Gregory Piatetsky-Shapiro, it is frequently cited as one of the top sources of data science news and influence by various industry watchers. Thomson Reuters: What are some use cases of data science that you find to be particularly valuable to organizations in this age of Big Data? GREGORY: Where people typically apply data science, probably not surprisingly, are in the areas of customer relationship management (CRM) and consumer analytics. Data science allows you to predict consumer behavior better and usually make incremental improvements and predictions, but those incremental improvements could translate to significant revenue.

Approximately one in nine men in the U.S. will develop prostate cancer in their lifetime, according to the National Cancer Institute, and more than 2.9 million patients diagnosed with it at some point are still alive today. And from a treatment perspective, it tends to be problematic -- prostate cancer is frequently non-aggressive, making it difficult to determine which, if any, procedures might be necessary. Google has made headway in diagnosing it, encouragingly, with the help of artificial intelligence (AI). In a paper ("Development and Validation of a Deep Learning Algorithm for Improving Gleason Scoring of Prostate Cancer") and accompanying blog post, Google AI researchers describe a system that uses the Gleason score -- a grading system that classifies cancer cells based on how closely they resemble normal prostate glands -- to detect problematic masses in samples. The goal, according to technical lead Martin Stumpe and Google AI Healthcare product manager Craig Mermel, was to develop AI that could perform Gleason grading objectively -- and precisely.