A gentle guided tour of Convolutional Neural Networks. Come lift the curtain and see how the magic is done.

A friendly explanation of how computers predict and generate sequences, based on Recurrent Neural Networks.

A novel approach for pitting algorithms against each other, called generative adversarial networks, or GANs, is showing promise for improving AI accuracy and for automatically generating objects that typically require human creativity. Yann LeCun, director of AI research at Facebook and professor at New York University, wrote that GANs and the variations now being proposed are "the most interesting idea in the last 10 years in machine learning." The reason for the excitement is the technique promises to automate the process of training algorithms in AI applications. The use of GANs makes it possible to improve the internal representation in AI algorithms for detecting important data features in the face of noise and other variations. In the long run, GANs could also help improve the algorithms used for discriminating fake objects from the real thing and improve results for detecting objects in poor images or speech in noisy environments.

Until now, the Machine Learning (ML) frameworks we've used at Borealis AI have varied according to individual preference. But as our applied team grows, we're finding that a preference-based system has certain shortcomings that have led to inefficiencies and delays in our research projects. As a result, we identified two main arguments in favour of standardizing a single framework for the lab. It has been our experience that independent frameworks do not often "play well" together. For example, a TensorFlow-based model applied to one research project would have to be rewritten in PyTorch for another project.

AI, machine learning, deep learning, autonomous systems and neural networks are not just buzzwords and phrases. Increased compute power, more efficient hardware and robust software, as well as an explosion in sensor data from the Internet of Things -- are fueling machine learning, and moving actionable data and intelligence towards edge devices. As AI makes devices, including smartphones and automobiles, more intelligent, mobile is becoming the key platform for enhancing all aspects of our lives, having an impact now and in the future.

Suppose you could merely imagine a computation, and a digital prostheses, an extension of your biological brain, would turn it into code that instantly realizes what you had in mind. Imagine looking at an image, dataset or set of equations and wanting to analyze and explore its meaning as an artistic whim or part of a scientific investigation. I don't mean you would use an existing software suite to produce a standard visualization, but rather you would make use of an extensive repository of existing code to assemble a new program analogous to how a composer draws upon a repertoire of musical motifs, themes and styles to construct new works, and tantamount to having a talented musical amanuensis who, in addition to copying your scores, takes liberties with your prior work, making small alterations here and there and occasionally adding new works of its own invention, novel but consistent with your taste and sensibilities. Perhaps the interaction would be wordless and you would express your objective by simply focusing your attention and guiding your imagination, the prostheses operating directly on patterns of activation arising in your primary sensory, proprioceptive and associative cortex that have become part of an extensive vocabulary that you now share with your personal digital amanuensis. Or perhaps it would involve a conversation conducted in subvocal, unarticulated speech in which you specify what it is you want to compute and your assistant asks questions to clarify your intention and the two of you share examples of input and output to ground your internal conversation in concrete terms. More than thirty years ago, Charles Rich and Richard Waters published an MIT AI Lab technical report [68] entitled The Programmer's Apprentice: A Research Overview. Whether they intended it or not, it would have been easy in those days for someone to misremember the title and inadvertently refer to it as "The Sorcerer's Apprentice" since computer programmers at the time were often characterized as wizards and most children were familiar with the Walt Disney movie Fantasia, featuring music written by Paul Dukas inspired by Goethe's poem of the same name

A main puzzle of deep neural networks (DNNs) revolves around the apparent absence of "overfitting", defined in this paper as follows: the expected error does not get worse when increasing the number of neurons or of iterations of gradient descent. This is surprising because of the large capacity demonstrated by DNNs to fit randomly labeled data and the absence of explicit regularization. Recent results by Srebro et al. provide a satisfying solution of the puzzle for linear networks used in binary classification. They prove that minimization of loss functions such as the logistic, the cross-entropy and the exp-loss yields asymptotic, "slow" convergence to the maximum margin solution for linearly separable datasets, independently of the initial conditions. Here we prove a similar result for nonlinear multilayer DNNs near zero minima of the empirical loss. The result holds for exponential-type losses but not for the square loss. In particular, we prove that the normalized weight matrix at each layer of a deep network converges to a minimum norm solution (in the separable case). Our analysis of the dynamical system corresponding to gradient descent of a multilayer network suggests a simple criterion for predicting the generalization performance of different zero minimizers of the empirical loss. This material is based upon work supported by the Center for Brains, Minds and Machines (CBMM), funded by NSF STC award CCF-1231216.

Applying deep learning methods to mammography assessment has remained a challenging topic. Dense noise with sparse expressions, mega-pixel raw data resolution, lack of diverse examples have all been factors affecting performance. The lack of pixel-level ground truths have especially limited segmentation methods in pushing beyond approximately bounding regions. We propose a classification approach grounded in high performance tissue assessment as an alternative to all-in-one localization and assessment models that is also capable of pinpointing the causal pixels. First, the objective of the mammography assessment task is formalized in the context of local tissue classifiers. Then, the accuracy of a convolutional neural net is evaluated on classifying patches of tissue with suspicious findings at varying scales, where highest obtained AUC is above $0.9$. The local evaluations of one such expert tissue classifier is used to augment the results of a heatmap regression model and additionally recover the exact causal regions at high resolution as a saliency image suitable for clinical settings.

While several approaches to face emotion recognition task are proposed in literature, none of them reports on power consumption nor inference time required to run the system in an embedded environment. Without adequate knowledge about these factors it is not clear whether we are actually able to provide accurate face emotion recognition in the embedded environment or not, and if not, how far we are from making it feasible and what are the biggest bottlenecks we face. The main goal of this paper is to answer these questions and to convey the message that instead of reporting only detection accuracy also power consumption and inference time should be reported as real usability of the proposed systems and their adoption in human computer interaction strongly depends on it. In this paper, we identify the state-of-the art face emotion recognition methods that are potentially suitable for embedded environment and the most frequently used datasets for this task. Our study shows that most of the performed experiments use datasets with posed expressions or in a particular experimental setup with special conditions for image collection. Since our goal is to evaluate the performance of the identified promising methods in the realistic scenario, we collect a new dataset with non-exaggerated emotions and we use it, in addition to the publicly available datasets, for the evaluation of detection accuracy, power consumption and inference time on three frequently used embedded devices with different computational capabilities. Our results show that gray images are still more suitable for embedded environment than color ones and that for most of the analyzed systems either inference time or energy consumption or both are limiting factor for their adoption in real-life embedded applications.

Despite the success of neural networks (NNs), there is still a concern among many over their "black box" nature. Why do they work? Here we present a simple analytic argument that NNs are in fact essentially polynomial regression models. This view will have various implications for NNs, e.g. providing an explanation for why convergence problems arise in NNs, and it gives rough guidance on avoiding overfitting. In addition, we use this phenomenon to predict and confirm a multicollinearity property of NNs not previously reported in the literature. Most importantly, given this loose correspondence, one may choose to routinely use polynomial models instead of NNs, thus avoiding some major problems of the latter, such as having to set many tuning parameters and dealing with convergence issues. We present a number of empirical results; in each case, the accuracy of the polynomial approach matches or exceeds that of NN approaches. A many-featured, open-source software package, polyreg, is available.