- neural connectivity maps of the brain-using the parallel execution of computer vision algorithms on high-performance compute clusters. These services and open-science data sets are publicly available at openconnecto.me. The system design inherits much from NoSQL scale-out and data-intensive computing architectures. We distribute data to cluster nodes by partitioning a spatial index. We direct I/O to different systems-reads to parallel disk arrays and writes to solid-state storage-to avoid I/O interference and maximize throughput. All programming interfaces are RESTful Web services, which are simple and stateless, improving scalability and usability. We include a performance evaluation of the production system, highlighting the effec-tiveness of spatial data organization.

Recent studies have shown that machine learning can improve the accuracy of detecting object boundaries in images. In the standard approach, a boundary detector is trained by minimizing its pixel-level disagreement with human boundary tracings. This naive metric is problematic because it is overly sensitive to boundary locations. This problem is solved by metrics provided with the Berkeley Segmentation Dataset, but these can be insensitive to topological differences, such as gaps in boundaries. Furthermore, the Berkeley metrics have not been useful as cost functions for supervised learning. Using concepts from digital topology, we propose a new metric called the warping error that tolerates disagreements over boundary location, penalizes topological disagreements, and can be used directly as a cost function for learning boundary detection, in a method that we call Boundary Learning by Optimization with Topological Constraints (BLOTC). We trained boundary detectors on electron microscopic images of neurons, using both BLOTC and standard training. BLOTC produced substantially better performance on a 1.2 million pixel test set, as measured by both the warping error and the Rand index evaluated on segmentations generated from the boundary labelings. We also find our approach yields significantly better segmentation performance than either gPb-OWT-UCM or multiscale normalized cut, as well as Boosted Edge Learning trained directly on our data.

It is still an unsolved problem how to optimally simulate self-motion using motion simulators. We investigated how a forward acceleration can be simulated as believably as possible on a hexapod motion platform equipped with a projection screen. Human participants rated the believability of brief forward accelerations. These were simulated as visual forward accelerations over a ground plane with people as size cues, presented together with brief forward surge translations and backward pitches of the platform, and synchronous random up--down movements of the camera in the visual scene and the platform. The magnitudes of all of the parameters were varied independently across trials. Even though variability between participants was high, the most believable simulation occurred when visual accelerations were combined with backward pitches of the platform, which changed the gravitoinertial vector direction approximately consistent with the visual acceleration. However, a wide range of platform pitches was accepted as believable. With high visual acceleration cues most participants reported trials as realistic even when the platform tilt rate was above vestibular canal thresholds reported in other works. Other manipulated parameters had only a mild influence on the responses. These results can be used to optimize motion-cueing algorithms for simulating linear accelerations in motion simulators.

No abstract available.