Abstract: COVID-19 had infected 127 million people and killed 2.5 million by March 31, 2021. COVID-19 patients and the highly contagious disease must be diagnosed quickly. Since non-contrast chest computed tomography (CT) is clinically useful…
Abstract: Click-through rate (CTR) prediction, which predicts the likelihood of a user clicking on an item, has become increasingly important in recommender systems. Deep learning models that automatically extract user interest from behavior have been…
Abstract: Meta-learning, or learning-to-learn, has exploded in popularity. Meta-learning uses multiple learning episodes to improve the learning algorithm, unlike traditional AI approaches that start from scratch. This paradigm can address data, computation, and generalization bottlenecks…
Abstract: Graph link prediction is a classic graph analytical problem with many applications. Current link prediction methods use deep learning to compute features from subgraphs centered at two neighboring nodes and predict the link label…
Abstract: Safety requires drowsiness detection. An automatic drowsiness detection system using pervasive sensor data (e.g., video, physiology) and machine learning has been extensively researched. However, most existing methods are based on complicated wearables (e.g., electroencephalogram)…
Abstract: Local key action regions can improve CNN-based visual action recognition. Self-attention requires focusing on important details and ignoring others. Self-attention aids action recognition. Current self-attention methods ignore local feature vector correlations at spatial positions…
Abstract: Annotating chest X-rays saves time and effort. Recent weakly-supervised algorithms like multi-instance learning (MIL) and class activation maps (CAM) attempted this task, but they often produce inaccurate or incomplete regions. One reason is the…
Abstract: Long-term image restoration and manipulation goals include learning a good image prior. Existing methods like deep image prior (DIP) capture low-level image statistics, but there are still gaps toward capturing rich image semantics like…
Abstract: Internet of things sensors are enabling data-driven traffic speed prediction, a foundation of advanced traffic management. However, current research studies mostly predict traffic for one hour ahead. Long-term prediction methods have error accumulation, exposure…
Abstract: BCI technology connects the brain to external devices. P300 wave extraction from electroencephalography (EEG) recordings is a popular BCI protocol. P300 waves are event-related potentials with a 300-ms latency after rare stimuli. This paper…
Abstract: Vehicle detection, which identifies vehicles as axis-aligned bounding boxes in still images, is widely used to estimate autonomous vehicle (AV) range, time-to-collision, and motion. Bounding boxes are convenient but too coarse to adapt to…
Abstract: Conventional deformable registration methods solve an image pair optimization model, which is computationally expensive. Recent deep learning-based methods estimate deformation quickly. These data-driven heuristic network architectures lack geometric constraints needed to generate plausible topology-preserving…
WhatsApp us