Search

The deployment of safe and trustworthy machine learning systems, and particularly complex black box neural networks, in real-world applications requires reliable and certified guarantees on their performance. The conformal prediction framework offers such formal guarantees by transforming any point into a set predictor with valid, finite-set, guarantees on the coverage of the true at a chosen level of confidence. Central to this methodology is the notion of the nonconformity score function that assigns to each example a measure of ''strangeness'' in comparison with the previously seen observations. While the coverage guarantees are maintained regardless of the nonconformity measure, the point predictor and the dataset, previous research has shown that the performance of a conformal model, as measured by its efficiency (the average size of the predicted sets) and its informativeness (the proportion of prediction sets that are singletons), is influenced by the choice of the nonconformity score function. The current work introduces the Penalized Inverse Probability (PIP) nonconformity score, and its regularized version RePIP, that allow the joint optimization of both efficiency and informativeness. Through toy examples and empirical results on the task of crop and weed image classification in agricultural robotics, the current work shows how PIP-based conformal classifiers exhibit precisely the desired behavior in comparison with other nonconformity measures and strike a good balance between informativeness and efficiency.

As deep learning predictive models become an integral part of a large spectrum of precision agricultural systems, a barrier to the adoption of such automated solutions is the lack of user trust in these highly complex, opaque and uncertain models. Indeed, deep neural networks are not equipped with any explicit guarantees that can be used to certify the system's performance, especially in highly varying uncontrolled environments such as the ones typically faced in computer vision for agriculture.Fortunately, certain methods developed in other communities can prove to be important for agricultural applications. This article presents the conformal prediction framework that provides valid statistical guarantees on the predictive performance of any black box prediction machine, with almost no assumptions, applied to the problem of deep visual classification of weeds and crops in real-world conditions. The framework is exposed with a focus on its practical aspects and special attention accorded to the Adaptive Prediction Sets (APS) approach that delivers marginal guarantees on the model's coverage. Marginal results are then shown to be insufficient to guarantee performance on all groups of individuals in the population as characterized by their environmental and pedo-climatic auxiliary data gathered during image acquisition.To tackle this shortcoming, group-conditional conformal approaches are presented: the ''classical'' method that consists of iteratively applying the APS procedure on all groups, and a proposed elegant reformulation and implementation of the procedure using quantile regression on group membership indicators. Empirical results showing the validity of the proposed approach are presented and compared to the marginal APS then discussed.

The present paper shows that warped Riemannian metrics, a class of Riemannian metrics which play a prominent role in Riemannian geometry, are also of fundamental importance in information geometry. Precisely, the paper features a new theorem, which states that the Rao-Fisher information metric of any location-scale model, defined on a Riemannian manifold, is a warped Riemannian metric, whenever this model is invariant under the action of some Lie group. This theorem is a valuable tool in finding the expression of the Rao-Fisher information metric of location-scale models defined on high-dimensional Riemannian manifolds. Indeed, a warped Riemannian metric is fully determined by only two functions of a single variable, irrespective of the dimension of the underlying Riemannian manifold. Starting from this theorem, several original contributions are made. The expression of the Rao-Fisher information metric of the Riemannian Gaussian model is provided, for the first time in the literature. A generalised definition of the Mahalanobis distance is introduced, which is applicable to any location-scale model defined on a Riemannian manifold. The solution of the geodesic equation is obtained, for any Rao-Fisher information metric defined in terms of warped Riemannian metrics. Finally, using a mixture of analytical and numerical computations, it is shown that the parameter space of the von Mises-Fisher model of $n$-dimensional directional data, when equipped with its Rao-Fisher information metric, becomes a Hadamard manifold, a simply-connected complete Riemannian manifold of negative sectional curvature, for $n = 2,\ldots,8$. Hopefully, in upcoming work, this will be proved for any value of $n$., Comment: first version, before submission

A novel efficient method for content-based image retrieval (CBIR) is developed in this paper using both texture and color features. Our motivation is to represent and characterize an input image by a set of local descriptors extracted at characteristic points (i.e. keypoints) within the image. Then, dissimilarity measure between images is calculated based on the geometric distance between the topological feature spaces (i.e. manifolds) formed by the sets of local descriptors generated from these images. In this work, we propose to extract and use the local extrema pixels as our feature points. Then, the so-called local extrema-based descriptor (LED) is generated for each keypoint by integrating all color, spatial as well as gradient information captured by a set of its nearest local extrema. Hence, each image is encoded by a LED feature point cloud and riemannian distances between these point clouds enable us to tackle CBIR. Experiments performed on Vistex, Stex and colored Brodatz texture databases using the proposed approach provide very efficient and competitive results compared to the state-of-the-art methods., Comment: This paper has been withdrawn by the author due to a crucial equation modification in part II.B

The Riemannian geometry of covariance matrices has been essential to several successful applications, in computer vision, biomedical signal and image processing, and radar data processing. For these applications, an important ongoing challenge is to develop Riemannian-geometric tools which are adapted to structured covariance matrices. The present paper proposes to meet this challenge by introducing a new class of probability distributions, Gaussian distributions of structured covariance matrices. These are Riemannian analogs of Gaussian distributions, which only sample from covariance matrices having a preassigned structure, such as complex, Toeplitz, or block-Toeplitz. The usefulness of these distributions stems from three features: (1) they are completely tractable, analytically or numerically, when dealing with large covariance matrices, (2) they provide a statistical foundation to the concept of structured Riemannian barycentre (i.e. Fr\'echet or geometric mean), (3) they lead to efficient statistical learning algorithms, which realise, among others, density estimation and classification of structured covariance matrices. The paper starts from the observation that several spaces of structured covariance matrices, considered from a geometric point of view, are Riemannian symmetric spaces. Accordingly, it develops an original theory of Gaussian distributions on Riemannian symmetric spaces, of their statistical inference, and of their relationship to the concept of Riemannian barycentre. Then, it uses this original theory to give a detailed description of Gaussian distributions of three kinds of structured covariance matrices, complex, Toeplitz, and block-Toeplitz. Finally, it describes algorithms for density estimation and classification of structured covariance matrices, based on Gaussian distribution mixture models., Comment: Final version

Data which lie in the space $\mathcal{P}_{m\,}$, of $m \times m$ symmetric positive definite matrices, (sometimes called tensor data), play a fundamental role in applications including medical imaging, computer vision, and radar signal processing. An open challenge, for these applications, is to find a class of probability distributions, which is able to capture the statistical properties of data in $\mathcal{P}_{m\,}$, as they arise in real-world situations. The present paper meets this challenge by introducing Riemannian Gaussian distributions on $\mathcal{P}_{m\,}$. Distributions of this kind were first considered by Pennec in $2006$. However, the present paper gives an exact expression of their probability density function for the first time in existing literature. This leads to two original contributions. First, a detailed study of statistical inference for Riemannian Gaussian distributions, uncovering the connection between maximum likelihood estimation and the concept of Riemannian centre of mass, widely used in applications. Second, the derivation and implementation of an expectation-maximisation algorithm, for the estimation of mixtures of Riemannian Gaussian distributions. The paper applies this new algorithm, to the classification of data in $\mathcal{P}_{m\,}$, (concretely, to the problem of texture classification, in computer vision), showing that it yields significantly better performance, in comparison to recent approaches., Comment: 21 pages, 1 table; accepted for publication in IEEE Trans Inf Theory