A Context-Enhanced De-identification System

Many modern entity recognition systems, including the current state-of-the-art de-identification systems, are based on bidirectional long short-term memory (biLSTM) units augmented by a conditional random field (CRF) sequence optimizer. These systems process the input sentence by sentence. This approach prevents the systems from capturing dependencies over sentence boundaries and makes accurate sentence boundary detection a prerequisite. Since sentence boundary detection can be problematic especially in clinical reports, where dependencies and co-references across sentence boundaries are abundant, these systems have clear limitations. In this study, we built a new system on the framework of one of the current state-of-the-art de-identification systems, NeuroNER, to overcome these limitations. This new system incorporates context embeddings through forward and backward n -grams without using sentence boundaries. Our context-enhanced de-identification (CEDI) system captures dependencies over sentence boundaries and bypasses the sentence boundary detection problem altogether. We enhanced this system with deep affix features and an attention mechanism to capture the pertinent parts of the input. The CEDI system outperforms NeuroNER on the 2006 i2b2 de-identification challenge dataset, the 2014 i2b2 shared task de-identification dataset, and the 2016 CEGS N-GRID de-identification dataset ( p < 0.01 ). All datasets comprise narrative clinical reports in English but contain different note types varying from discharge summaries to psychiatric notes. Enhancing CEDI with deep affix features and the attention mechanism further increased performance.

AI-Based Video Clipping of Soccer Events

The current gold standard for extracting highlight clips from soccer games is the use of manual annotations and clippings, where human operators define the start and end of an event and trim away the unwanted scenes. This is a tedious, time-consuming, and expensive task, to the extent of being rendered infeasible for use in lower league games. In this paper, we aim to automate the process of highlight generation using logo transition detection, scene boundary detection, and optional scene removal. We experiment with various approaches, using different neural network architectures on different datasets, and present two models that automatically find the appropriate time interval for extracting goal events. These models are evaluated both quantitatively and qualitatively, and the results show that we can detect logo and scene transitions with high accuracy and generate highlight clips that are highly acceptable for viewers. We conclude that there is considerable potential in automating the overall soccer video clipping process.

SVA-SSD: saliency visual attention single shot detector for building detection in low contrast high-resolution satellite images

Building detection in high-resolution satellite images has received great attention, as it is important to increase the accuracy of urban planning. The building boundary detection in the desert environment is a real challenge due to the nature of low contrast images in the desert environment. The traditional computer vision algorithms for building boundary detection lack scalability, robustness, and accuracy. On the other hand, deep learning detection algorithms have not been applied to such low contrast satellite images. So, there is a real need to employ deep learning algorithms for building detection tasks in low contrast high-resolution images. In this paper, we propose a novel building detection method based on a single-shot multi-box (SSD) detector. We develop the state-of-the-art SSD detection algorithm based on three approaches. First, we propose data-augmentation techniques to overcome the low contrast images’ appearance. Second, we develop the SSD backbone using a novel saliency visual attention mechanism. Moreover, we investigate several pre-trained networks performance and several fusion functions to increase the performance of the SSD backbone. The third approach is based on optimizing the anchor-boxes sizes which are used in the detection stage to increase the performance of the SSD head. During our experiments, we have prepared a new dataset for buildings inside Riyadh City, Saudi Arabia that consists of 3878 buildings. We have compared our proposed approach vs other approaches in the literature. The proposed system has achieved the highest average precision, recall, F1-score, and IOU performance. Our proposed method has achieved a fast average prediction time with the lowest variance for our testing set. Our experimental results are very promising and can be generalized to other object detection tasks in low contrast images.

Ego-Lane Index Estimation Based on Lane-Level Map and LiDAR Road Boundary Detection

Correct ego-lane index estimation is essential for lane change and decision making for intelligent vehicles, especially in global navigation satellite system (GNSS)-challenged environments. To achieve this, we propose an ego-lane index estimation approach in an urban scenario based on particle filter (PF). The particles are initialized and propagated by dead reckoning with inertial measurement unit (IMU) and odometry. A lane-level map is used to navigate the particles taking advantage of topologic and geometric information of lanes. GNSS single-point positioning (SPP) can provide position estimation with meter-level accuracy in urban environments, which can limit drift introduced by dead reckoning for updating the weight of each particle. Light detection and ranging (LiDAR) is a common sensor in an intelligent vehicle. A LiDAR-based road boundary detection method provides distance measurements from the vehicle to the left/right road boundaries, which provides a measurement for importance weighting. However, the high precision of the LiDAR measurements may put a tight constraint on the distribution of particles, which can lead to particle degeneration with sparse particle sets. To deal with this problem, we propose a novel step that shifts particles laterally based on LiDAR measurements instead of importance weighting in the traditional PF scheme. We tested our methods on an urban expressway at a low traffic volume period to ensure road boundaries can be detected by LiDAR measurements at most time steps. Experimental results prove that our improved PF scheme can correctly estimate ego-lane index at all time steps, while the traditional PF scheme produces wrong estimations at some time steps.

The Network Structure of Occupations: Fragmentation, Differentiation, and Contagion

Occupational structure is commonly viewed as either hierarchical or organized around stable classes. Yet, recent studies have proposed to describe occupational structure as a network, where the mobility of workers demarcates boundaries. Moving beyond boundary detection, this article develops occupational network as a dynamic system in which between-occupation exchange is shaped by occupational similarities, and occupational attributes are in turn responsive to mobility patterns. We illustrate this perspective with the exchange networks of detailed occupations. Our analysis shows that the U.S. occupational structure has become more fragmented. The division was in part associated with the emerging importance of age composition, as well as those of quantitative, creative, and social tasks. The fragmentation reduced wage contagion and therefore contributed to a greater between-occupation wage dispersion. These results indicate that occupational attributes and mobility are co-constitutive, and that a network perspective provides a unifying framework for the study of stratification and mobility.