Discriminative learning effectively predicts true object class for image classification. However, it often results in false positives for outliers, posing critical concerns in applications like autonomous driving and video surveillance systems. Previous attempts to address this challenge involved training image classifiers through contrastive learning using actual outlier data or synthesizing outliers for self-supervised learning. Furthermore, unsupervised generative modeling of inliers in pixel space has shown limited success for outlier detection. In this work, we introduce a quantile-based maximum likelihood objective for learning the inlier distribution to improve the outlier separation during inference. Our approach fits a normalizing flow to pre-trained discriminative features and detects the outliers according to the evaluated log-likelihood. The experimental evaluation demonstrates the effectiveness of our method as it surpasses the performance of the state-of-the-art unsupervised methods for outlier detection. The results are also competitive compared with a recent self-supervised approach for outlier detection. Our work allows to reduce dependency on well-sampled negative training data, which is especially important for domains like medical diagnostics or remote sensing.

Real-world deployment of reliable object detectors is crucial for applications such as autonomous driving. However, general-purpose object detectors like Faster R-CNN are prone to providing overconfident predictions for outlier objects. Recent outlier-aware object detection approaches estimate the density of instance-wide features with class-conditional Gaussians and train on synthesized outlier features from their low-likelihood regions. However, this strategy does not guarantee that the synthesized outlier features will have a low likelihood according to the other class-conditional Gaussians. We propose a novel outlier-aware object detection framework that distinguishes outliers from inlier objects by learning the joint data distribution of all inlier classes with an invertible normalizing flow. The appropriate sampling of the flow model ensures that the synthesized outliers have a lower likelihood than inliers of all object classes, thereby modeling a better decision boundary between inlier and outlier objects. Our approach significantly outperforms the state-of-the-art for outlier-aware object detection on both image and video datasets.

Deep learning techniques are recently being used in fundus image analysis and diabetic retinopathy detection. Microaneurysms are important indicators of diabetic retinopathy progression. The authors introduce a two‐stage deep learning approach for microaneurysms segmentation using multiple scales of the input with selective sampling and embedding triplet loss. The proposed approach facilitates a region proposal fully convolutional neural network trained on segmented patches and a patch‐wise refinement network for improving the results suggested by the first stage hypothesis. To enhance the discriminative power of the second stage refinement network, the authors use triplet embedding loss with a selective sampling routine that dynamically assigns sampling probabilities to the oversampled class patches. This approach introduces a relative improvement over the vanilla fully convolutional neural network on the Indian Diabetic Retinopathy Image Data set segmentation data set. The proposed segmentation is incorporated in a classification model to solve two downstream tasks for diabetic retinopathy detection and referable diabetic retinopathy detection. The classification tasks are trained on the Kaggle diabetic retinopathy challenge data set and evaluated on the Messidor data. The authors show that adding the segmentation enhances the classification performance and achieves comparable performance to the state‐of‐the‐art models.

Machine learning and especially deep learning techniques are dominating medical image and data analysis. This article reviews machine learning approaches proposed for diagnosing ophthalmic diseases during the last four years. Three diseases are addressed in this survey, namely diabetic retinopathy, age-related macular degeneration, and glaucoma. The review covers over 60 publications and 25 public datasets and challenges related to the detection, grading, and lesion segmentation of the three considered diseases. Each section provides a summary of the public datasets and challenges related to each pathology and the current methods that have been applied to the problem. Furthermore, the recent machine learning approaches used for retinal vessels segmentation, and methods of retinal layers and fluid segmentation are reviewed. Two main imaging modalities are considered in this survey, namely color fundus imaging, and optical coherence tomography. Machine learning approaches that use eye measurements and visual field data for glaucoma detection are also included in the survey. Finally, the authors provide their views, expectations and the limitations of the future of these techniques in the clinical practice.

Subretinal injection is a delicate and complex microsurgery, which requires surgeons to inject the therapeutic substance in a pre-operatively defined and intra-operatively updated subretinal target area. Due to the lack of subretinal visual feedback, it is hard to sense the insertion depth during the procedure, thus affecting the results of surgical outcome and hindering the widespread use of this treatment. This paper presents a novel approach to estimate the 3D position of the needle under the retina using the information from microscope-integrated Intraoperative Optical Coherence Tomography (iOCT). We evaluated our approach on both tissue phantom and ex-vivo porcine eyes. Evaluation results show that the average error in distance measurement is 4.7 μm (maximum of 16.5 μm). We furthermore, verified the feasibility of the proposed method to track the insertion depth of needle in robot-assisted subretinal injection.

Sonic interaction as a technique for conveying information has advantages over conventional visual augmented reality methods specially when augmenting the visual field with extra information brings distraction. Sonification of knowledge extracted by applying computational methods to sensory data is a well-established concept. However, some aspects of sonic interaction design such as aesthetics, the cognitive effort required for perceiving information, and avoiding alarm fatigue are not well studied in literature. In this work, we present a sonification scheme based on employment of physical modeling sound synthesis which targets focus demanding tasks requiring extreme precision. Proposed mapping techniques are designed to require minimum training for users to adapt to and minimum mental effort to interpret the conveyed information. Two experiments are conducted to assess the feasibility of the proposed method and compare it against visual augmented reality in high precision tasks. The observed quantitative results suggest that utilizing sound patches generated by physical modeling achieve the desired goal of improving the user experience and general task performance with minimal training.

Augmented Reality (AR) in microscopic surgery has been subject of several studies in the past two decades. Nevertheless, AR has not found its way into everyday microsurgical workflows. The introduction of new surgical microscopes equipped with Optical Coherence Tomography (OCT) enables the surgeons to perform multimodal (optical and OCT) imaging in the operating room. Taking full advantage of such elaborate source of information requires sophisticated intraoperative image fusion, information extraction, guidance and visualization methods. Medical AR is a unique approach to facilitate utilization of multimodal medical imaging devices. Here we propose a novel medical AR solution to the long-known problem of determining the distance between the surgical instrument tip and the underlying tissue in ophthalmic surgery to further pave the way of AR into the surgical theater. Our method brings augmented reality to OCT for the first time by augmenting the surgeon's view of the OCT images with an estimated instrument cross-section shape and distance to the retinal surface using only information from the shadow of the instrument in intraoperative OCT images. We demonstrate the applicability of our method in retinal surgery using a phantom eye and evaluate the accuracy of the augmented information using a micromanipulator.

In this paper, we present a fusion algorithm supplemented with appropriate visualization by selecting relevant information from different modalities in mixed and augmented reality (AR). This encompasses a learning based method upon relevance of information, defined by an expert, which ultimately enables confident interventional decisions based on mixed reality (MR) images. The performance of our developed fusion and tailored visualization techniques was evaluated by employing X-ray/optical images during surgery and validated qualitatively using a 5-point Likert scale. Our observations indicated that the proposed technique provided semantic contextual information about underlying pixels and in general was preferred over the traditional pixel-wise linear alpha-blending method.