Accurate diagnosis of psychiatric disorders plays a critical role in improving quality of life for patients and potentially supports the development of new treatments. Many studies have been conducted on machine learning techniques that seek brain imaging data for specific biomarkers of disorders. These studies have encountered the following dilemma: An end-to-end classification overfits to a small number of high-dimensional samples but unsupervised feature-extraction has the risk of extracting a signal of no interest. In addition, such studies often provided only diagnoses for patients without presenting the reasons for these diagnoses. This study proposed a deep neural generative model of resting-state functional magnetic resonance imaging (fMRI) data. The proposed model is conditioned by the assumption of the subject's state and estimates the posterior probability of the subject's state given the imaging data, using Bayes' rule. This study applied the proposed model to diagnose schizophrenia and bipolar disorders. Diagnosis accuracy was improved by a large margin over competitive approaches, namely a support vector machine, logistic regression, and multilayer perceptron with or without unsupervised feature-extractors in addition to a Gaussian mixture model. The proposed model visualizes brain regions largely related to the disorders, thus motivating further biological investigation.

We propose a recurrent extension of the Ladder networks whose structure is motivated by the inference required in hierarchical latent variable models. We demonstrate that the recurrent Ladder is able to handle a wide variety of complex learning tasks that benefit from iterative inference and temporal modeling. The architecture shows close-to-optimal results on temporal modeling of video data, competitive results on music modeling, and improved perceptual grouping based on higher order abstractions, such as stochastic textures and motion cues. We present results for fully supervised, semi-supervised, and unsupervised tasks. The results suggest that the proposed architecture and principles are powerful tools for learning a hierarchy of abstractions, learning iterative inference and handling temporal information.

Geometric variations of objects, which do not modify the object class, pose a major challenge for object recognition. These variations could be rigid as well as non-rigid transformations. In this paper, we design a framework for training deformable classifiers, where latent transformation variables are introduced, and a transformation of the object image to a reference instantiation is computed in terms of the classifier output, separately for each class. The classifier outputs for each class, after transformation, are compared to yield the final decision. As a by-product of the classification this yields a transformation of the input object to a reference pose, which can be used for downstream tasks such as the computation of object support. We apply a two-step training mechanism for our framework, which alternates between optimizing over the latent transformation variables and the classifier parameters to minimize the loss function. We show that multilayer perceptrons, also known as deep networks, are well suited for this approach and achieve state of the art results on the rotated MNIST and the Google Earth dataset, and produce competitive results on MNIST and CIFAR-10 when training on smaller subsets of training data.

Digital pathology is not only one of the most promising fields of diagnostic medicine, but at the same time a hot topic for fundamental research. Digital pathology is not just the transfer of histopathological slides into digital representations. The combination of different data sources (images, patient records, and *omics data) together with current advances in artificial intelligence/machine learning enable to make novel information accessible and quantifiable to a human expert, which is not yet available and not exploited in current medical settings. The grand goal is to reach a level of usable intelligence to understand the data in the context of an application task, thereby making machine decisions transparent, interpretable and explainable. The foundation of such an "augmented pathologist" needs an integrated approach: While machine learning algorithms require many thousands of training examples, a human expert is often confronted with only a few data points. Interestingly, humans can learn from such few examples and are able to instantly interpret complex patterns. Consequently, the grand goal is to combine the possibilities of artificial intelligence with human intelligence and to find a well-suited balance between them to enable what neither of them could do on their own. This can raise the quality of education, diagnosis, prognosis and prediction of cancer and other diseases. In this paper we describe some (incomplete) research issues which we believe should be addressed in an integrated and concerted effort for paving the way towards the augmented pathologist.

In this work, we show the intrinsic relations between optimal transportation and convex geometry, especially the variational approach to solve Alexandrov problem: constructing a convex polytope with prescribed face normals and volumes. This leads to a geometric interpretation to generative models, and leads to a novel framework for generative models. By using the optimal transportation view of GAN model, we show that the discriminator computes the Kantorovich potential, the generator calculates the transportation map. For a large class of transportation costs, the Kantorovich potential can give the optimal transportation map by a close-form formula. Therefore, it is sufficient to solely optimize the discriminator. This shows the adversarial competition can be avoided, and the computational architecture can be simplified. Preliminary experimental results show the geometric method outperforms WGAN for approximating probability measures with multiple clusters in low dimensional space.

We introduce a general-purpose conditioning method for neural networks called FiLM: Feature-wise Linear Modulation. FiLM layers influence neural network computation via a simple, feature-wise affine transformation based on conditioning information. We show that FiLM layers are highly effective for visual reasoning - answering image-related questions which require a multi-step, high-level process - a task which has proven difficult for standard deep learning methods that do not explicitly model reasoning. Specifically, we show on visual reasoning tasks that FiLM layers 1) halve state-of-the-art error for the CLEVR benchmark, 2) modulate features in a coherent manner, 3) are robust to ablations and architectural modifications, and 4) generalize well to challenging, new data from few examples or even zero-shot.

Achieving artificial visual reasoning -- the ability to answer image-related questions which require a multi-step, high-level process -- is an important step towards artificial general intelligence. This multi-modal task requires learning a questiondependent, structured reasoning process over images from language. Standard deep learning approaches tend to exploit biases in the data rather than learn this underlying structure, while leading methods learn to visually reason successfully but are handcrafted for reasoning. We show that a general-purpose, Conditional Batch Normalization approach achieves state-ofthe-art results on the CLEVR Visual Reasoning benchmark with a 2.4% error rate. We outperform the next best end-to-end method (4.5%) and even methods that use extra supervision (3.1%). We probe our model to shed light on how it reasons, showing it has learned a question-dependent, multi-step process. Previous work has operated under the assumption that visual reasoning calls for a specialized architecture, but we show that a general architecture with proper conditioning can learn to visually reason effectively. Index Terms: Deep Learning, Language and Vision Note: A full paper extending this study is available at http: //arxiv.org/abs/1709.07871, with additional references, experiments, and analysis.

While deep learning has had significant successes in computer vision thanks to the abundance of visual data, collecting sufficiently large real-world datasets for robot learning can be costly. To increase the practicality of these techniques on real robots, we propose a modular deep reinforcement learning method capable of transferring models trained in simulation to a real-world robotic task. We introduce a bottleneck between perception and control, enabling the networks to be trained independently, but then merged and fine-tuned in an end-to-end manner to further improve hand-eye coordination. On a canonical, planar visually-guided robot reaching task a fine-tuned accuracy of 1.6 pixels is achieved, a significant improvement over naive transfer (17.5 pixels), showing the potential for more complicated and broader applications. Our method provides a technique for more efficient learning and transfer of visuo-motor policies for real robotic systems without relying entirely on large real-world robot datasets.

With global climate change, devastating hurricanes are occurring with higher frequency. After a hurricane, roads are often flooded or washed out, making them treacherous for motorists. According to The Weather Channel, almost two of every three U.S. flash flood deaths from 1995–2010, excluding fatalities from Hurricane Katrina, occurred in vehicles. During my Insight A.I. Fellowship, I designed a system that detects flooded roads and created an interactive map app. Using state of the art computer vision deep learning methods, the system automatically annotates flooded, washed out, or otherwise severely damaged roads from satellite imagery.

Artificial intelligence and deep learning may soon deliver 24/7 advisors that double as insurance fraud detectors. First, insurers must learn to use data. Stéphane Tremblay, team leader at the National Research Council's (NRC) Data Analytics Centre, specializes in automated learning. The Analytics Centre helps businesses in all sectors face the challenge of big data. The NRC invests $1 billion in research and development each year.