We present an interpretable framework for path prediction that leverages dependencies between agents' behaviors and their spatial navigation environment. We exploit two sources of information: the past motion trajectory of the agent of interest and a wide top-view image of the navigation scene. We propose a Clairvoyant Attentive Recurrent Network (CAR-Net) that learns where to look in a large image of the scene when solving the path prediction task. Our method can attend to any area, or combination of areas, within the raw image (e.g., road intersections) when predicting the trajectory of the agent. This allows us to visualize fine-grained semantic elements of navigation scenes that influence the prediction of trajectories. To study the impact of space on agents' trajectories, we build a new dataset made of top-view images of hundreds of scenes (Formula One racing tracks) where agents' behaviors are heavily influenced by known areas in the images (e.g., upcoming turns). CAR-Net successfully attends to these salient regions. Additionally, CAR-Net reaches state-of-the-art accuracy on the standard trajectory forecasting benchmark, Stanford Drone Dataset (SDD). Finally, we show CAR-Net's ability to generalize to unseen scenes.

We develop an algorithm that accurately detects Atrial Fibrillation (AF) episodes from photoplethysmograms (PPG) recorded in ambulatory free-living conditions. We collect and annotate a dataset containing more than 4000 hours of PPG recorded from a wrist-worn device. Using a 50-layer convolutional neural network, we achieve a test AUC of 95% and show robustness to motion artifacts inherent to PPG signals. Continuous and accurate detection of AF from PPG has the potential to transform consumer wearable devices into clinically useful medical monitoring tools. * indicates equal contribution arXiv:1811.07774v2 [physics.med-ph]

Access to quantitative, robust, yet affordable diagnostic tools is necessary to reduce global infectious disease burden. Manual microscopy has served as a bedrock for diagnostics with wide adaptability, although at a cost of tedious labor and human errors. Automated robotic microscopes are poised to enable a new era of smart field microscopy but current platforms remain cost prohibitive and largely inflexible, especially for resource poor and field settings. Here we present Octopi, a low-cost ($250-$500) and reconfigurable autonomous microscopy platform capable of automated slide scanning and correlated bright-field and fluorescence imaging. Being highly modular, it also provides a framework for new disease-specific modules to be developed.We demonstrate the power of the platform by applying it to automated detection of malaria parasites in blood smears. Specifically, we discovered a spectral shift on the order of 10 nm for DAPI-stained Plasmodium falciparum malaria parasites. This shift allowed us to detect the parasites with a low magnification (equivalent to 10x) large field of view (2.56 mm 2 ) module.Combined with automated slide scanning, real time computer vision and machine learningbased classification, Octopi is able to screen more than 1.5 million red blood cells per minute for parasitemia quantification, with estimated diagnostic sensitivity and specificity exceeding 90% at parasitemia of 50/ul and 100% for parasitemia higher than 150/l. With different modules, we further showed imaging of tissue slice and sputum sample on the platform. With roughly two orders of magnitude in cost reduction, Octopi opens up the possibility of a large robotic microscope network for improved disease diagnosis while providing an avenue for collective efforts for development of modular instruments.One sentence summary: We developed a low-cost ($250-$500) automated imaging platform that can quantify malaria parasitemia by scanning 1.5 million red blood cells per minute.Lack of cost-effective diagnostics is a major hurdle in global fight against infectious disease, 2 specially in resource poor settings [1]. This leaves our world in a highly vulnerable position 3 with therapeutic drugs being either overused, leading to drug resistant strains or not acces-4 sible to people who actually need these treatments [2]. Since health care is delivered around 5 the world in a tiered structure, local context such as high cost, lack of trained personal or 6 low throughput of many available diagnostics tests plays a large detrimental role on quality 7 of delivered health-care [3]. 8 Because of the versatility and wide adoption of manual microscopy [4] and its role in 9 direct visual identification of parasites [5], it remains a WHO gold standard for numerous 10 diseases [1]. Despite technological advancements in related fields, the practice of conventional 11 manual microscopy has remained largely unchanged over the last half century and suffers 12 from several drawbacks. With an average lab technician spending 6 to 8 hours imaging and 13 ...

No abstract

No abstract