Most segmentation methods in child brain MRI are supervised and are based on global intensity distributions of major brain structures. The successful implementation of a supervised approach depends on availability of an age-appropriate probabilistic brain atlas. For the study of early normal brain development, the construction of such a brain atlas remains a significant challenge. Moreover, using global intensity statistics leads to inaccurate detection of major brain tissue classes due to substantial intensity variations of MR signal within the constituent parts of early developing brain. In order to overcome these methodological limitations we develop a local, semi-supervised framework. It is based on Kernel Fisher Discriminant Analysis (KFDA) for pattern recognition, combined with an objective structural similarity index (SSIM) for perceptual image quality assessment. The proposed method performs optimal brain partitioning into subdomains having different average intensity values followed by SSIM-guided computation of separating surfaces between the constituent brain parts. The classified image subdomains are then stitched slice by slice via simulated annealing to form a global image of the classified brain. In this paper, we consider classification into major tissue classes (white matter and grey matter) and the cerebrospinal fluid and illustrate the proposed framework on examples of brain templates for ages 8 to 11 months and ages 44 to 60 months. We show that our method improves detection of the tissue classes by its comparison to state-of-the-art classification techniques known as Partial Volume Estimation.

Learning a Gaussian Mixture Model (GMM) is hard when the number of parameters is too large given the amount of available data. As a remedy, we propose restricting the GMM to a Gaussian Markov Random Field Mixture Model (GMRF-MM), as well as a new method for estimating the latter's sparse precision (i.e., inverse covariance) matrices. When the sparsity pattern of each matrix is known, we propose an efficient optimization method for the Maximum Likelihood Estimate (MLE) of that matrix. When it is unknown, we utilize the popular Graphical LASSO (GLASSO) to estimate that pattern. However, we show that even for a single Gaussian, when GLASSO is tuned to successfully estimate the sparsity pattern, it does so at the price of a substantial bias of the values of the nonzero entries of the matrix, and we show that this problem only worsens in a mixture setting. To overcome this, we discard the non-zero values estimated by GLASSO, keep only its pattern estimate and use it within the proposed MLE method. This yields an effective two-step procedure that removes the bias. We show that our "debiasing" approach outperforms GLASSO in both the single-GMRF and the GMRF-MM cases. We also show that when learning priors for image patches, our method outperforms GLASSO even if we merely use an educated guess about the sparsity pattern, and that our GMRF-MM outperforms the baseline GMM on real and synthetic high-dimensional datasets. Our code is available at \url{https://github.com/shahaffind/GMRF-MM}.

Moog Inc. has automated the evaluation of copper (Cu) alloy grain size using a deep-learning convolutional neural network (CNN). The proof-of-concept automated image acquisition and batch-wise image processing offers the potential for significantly reduced labor, improved accuracy of grain evaluation, and decreased overall turnaround times for approving Cu alloy bar stock for use in flight critical aircraft hardware. A classification accuracy of 91.1% on individual sub-images of the Cu alloy coupons was achieved. Process development included minimizing the variation in acquired image color, brightness, and resolution to create a dataset with 12300 sub-images, and then optimizing the CNN hyperparameters on this dataset using statistical design of experiments (DoE). Over the development of the automated Cu alloy grain size evaluation, a degree of "explainability" in the artificial intelligence (XAI) output was realized, based on the decomposition of the large raw images into many smaller dataset sub-images, through the ability to explain the CNN ensemble image output via inspection of the classification results from the individual smaller sub-images.

Regularization-based approaches for injecting constraints in Machine Learning (ML) were introduced (see e.g. Given the recent interest in ethical and trustworthy AI, however, several works are resorting to these approaches for enforcing desired properties over a ML model (e.g. The regularization function C denotes a vector of (nonnegative) constraint violation indices for m constraints, while λ 0 is a vector of weights (or multipliers). As an example, in a regression problem we may desire a specific output ordering for two input vectors in the training set. If n is even, the term is 0 for perfectly balanced classifications.

If science was a journey, then its destination would be the discovery of simple explanations to complex phenomena. There was a time when the existence of tides, the planet's orbit around the sun, and the observation that "things fall down" were all largely considered to be independent phenomena--until 1687, when Isaac Newton formulated his law of gravitation that provided an elegantly simple explanation to all of these (and many more). Physics has made tremendous progress over the last few centuries, but the thriving field of deep learning is still very much at the beginning of its journey--often lacking a detailed understanding of the underlying principles. For some time, the tremendous success of deep learning has perhaps overshadowed the need to thoroughly understand the behaviour of Deep Neural Networks (DNNs). In an ever-increasing pace, DNNs were reported as having achieved human-level object classification performance [1], beating world-class human Go, Poker, and Starcraft players [2, 3], detecting cancer from X-ray scans [4], translating text across languages [5], helping combat climate change [6], and accelerating the pace of scientific progress itself [7]. Because of these successes, deep learning has gained a strong influence on our lives and society.

We present an interpretable deep model for fine-grained visual recognition. At the core of our method lies the integration of region-based part discovery and attribution within a deep neural network. Our model is trained using image-level object labels, and provides an interpretation of its results via the segmentation of object parts and the identification of their contributions towards classification. To facilitate the learning of object parts without direct supervision, we explore a simple prior of the occurrence of object parts. We demonstrate that this prior, when combined with our region-based part discovery and attribution, leads to an interpretable model that remains highly accurate. Our model is evaluated on major fine-grained recognition datasets, including CUB-200, CelebA and iNaturalist. Our results compare favorably to state-of-the-art methods on classification tasks, and our method outperforms previous approaches on the localization of object parts.

We consider a deep structured linear network under sparsity constraints. We study sharp conditions guaranteeing the stability of the optimal parameters defining the network. More precisely, we provide sharp conditions on the network architecture and the sample under which the error on the parameters defining the network scales linearly with the reconstruction error (i.e. the risk). Therefore, under these conditions, the weights obtained with a successful algorithms are well defined and only depend on the architecture of the network and the sample. The features in the latent spaces are stably defined. The stability property is required in order to interpret the features defined in the latent spaces. It can also lead to a guarantee on the statistical risk. This is what motivates this study. The analysis is based on the recently proposed Tensorial Lifting. The particularity of this paper is to consider a sparsity prior. This leads to a better stability constant. As an illustration, we detail the analysis and provide sharp stability guarantees for convolutional linear network under sparsity prior. In this analysis, we distinguish the role of the network architecture and the sample input. This highlights the requirements on the data in connection to parameter stability.

Automated computer vision systems have been applied in many domains including security, law enforcement, and personal devices, but recent reports suggest that these systems may produce biased results, discriminating against people in certain demographic groups. Diagnosing and understanding the underlying true causes of model biases, however, are challenging tasks because modern computer vision systems rely on complex black-box models whose behaviors are hard to decode. We propose to use an encoder-decoder network developed for image attribute manipulation to synthesize facial images varying in the dimensions of gender and race while keeping other signals intact. We use these synthesized images to measure counterfactual fairness of commercial computer vision classifiers by examining the degree to which these classifiers are affected by gender and racial cues controlled in the images, e.g., feminine faces may elicit higher scores for the concept of nurse and lower scores for STEM-related concepts. We also report the skewed gender representations in an online search service on profession-related keywords, which may explain the origin of the biases encoded in the models.

We argue that while this discourse on data ethics is of critical importance, it is missing one fundamental point: If more and more efforts in business, government, science, and our daily lives are data-driven, we should pay more attention to what exactly we are driven by. Therefore, we need more debate on what fundamental properties constitute data. In the first section of the paper, we work from the fundamental properties necessary for statistical computation to a definition of statistical data. We define a statistical datum as the coming together of substantive and numerical properties and differentiate between qualitative and quantitative data. Subsequently, we qualify our definition by arguing that for data to be practically useful, it needs to be commensurable in a manner that reveals meaningful differences that allow for the generation of relevant insights through statistical methodologies. In the second section, we focus on what our conception of data can contribute to the discourse on data ethics and beyond. First, we hold that the need for useful data to be commensurable rules out an understanding of properties as fundamentally unique or equal. Second, we argue that practical concerns lead us to increasingly standardize how we operationalize a substantive property; in other words, how we formalize the relationship between the substantive and numerical properties of data. Thereby, we also standardize the interpretation of a property. With our increasing reliance on data and data technologies, these two characteristics of data affect our collective conception of reality. Statistical data's exclusion of the fundamentally unique and equal influences our perspective on the world, and the standardization of substantive properties can be viewed as profound ontological practice, entrenching ever more pervasive interpretations of phenomena in our everyday lives.

The human brain provides a range of functions such as expressing emotions, controlling the rate of breathing, etc., and its study has attracted the interest of scientists for many years. As machine learning models become more sophisticated, and bio-metric data becomes more readily available through new non-invasive technologies, it becomes increasingly possible to gain access to interesting biometric data that could revolutionize Human-Computer Interaction. In this research, we propose a method to assess and quantify human attention levels and their effects on learning. In our study, we employ a brain computer interface (BCI) capable of detecting brain wave activity and displaying the corresponding electroencephalograms (EEG). We train recurrent neural networks (RNNS) to identify the type of activity an individual is performing.