Main Menu (Mobile)- Block

Main Menu - Block

custom | custom

Search Results

filters_region_cap | custom

Filter

facetapi-Q2b17qCsTdECvJIqZJgYMaGsr8vANl1n | block
facetapi-W9JlIB1X0bjs93n1Alu3wHJQTTgDCBGe | block

Associated Project Team

facetapi-61yz1V0li8B1bixrCWxdAe2aYiEXdhd0 | block
facetapi-PV5lg7xuz68EAY8eakJzrcmwtdGEnxR0 | block
general_search_page-panel_pane_1 | views_panes

11 Janelia Publications

Showing 1-10 of 11 results
Your Criteria:
    12/28/14 | Neural computation for rehabilitation.
    Hu X, Wang Y, Zhao T, Gunduz A
    Biomed Research International. 2014;2014:603985. doi: 10.1155/2014/603985
    11/03/14 | Protecting integrated circuits from piracy with test-aware logic locking.
    Plaza SM, Markov IL
    ICCAD '14 Proceedings of the 2014 IEEE/ACM International Conference on Computer-Aided Design. 2014 Nov 03:262-269. doi: 10.1109/ICCAD.2014.7001361

    The increasing IC manufacturing cost encourages a business model where design houses outsource IC fabrication to remote foundries. Despite cost savings, this model exposes design houses to IC piracy as remote foundries can manufacture in excess to sell on the black market. Recent efforts in digital hardware security aim to thwart piracy by using XOR-based chip locking, cryptography, and active metering. To counter direct attacks and lower the exposure of unlocked circuits to the foundry, we introduce a multiplexor-based locking strategy that preserves test response allowing IC testing by an untrusted party before activation. We demonstrate a simple yet effective attack against a locked circuit that does not preserve test response, and validate the effectiveness of our locking strategy on IWLS 2005 benchmarks.

    View Publication Page
    09/14/14 | Small sample learning of superpixel classifiers for EM segmentation.
    Parag T, Plaza S, Scheffer L
    Medical Image Computing and Computer-Assisted Intervention. 2014;17(Pt 1):389-97

    Pixel and superpixel classifiers have become essential tools for EM segmentation algorithms. Training these classifiers remains a major bottleneck primarily due to the requirement of completely annotating the dataset which is tedious, error-prone and costly. In this paper, we propose an interactive learning scheme for the superpixel classifier for EM segmentation. Our algorithm is 'active semi-supervised' because it requests the labels of a small number of examples from user and applies label propagation technique to generate these queries. Using only a small set (< 20%) of all datapoints, the proposed algorithm consistently generates a classifier almost as accurate as that estimated from a complete groundtruth. We provide segmentation results on multiple datasets to show the strength of these classifiers.

    View Publication Page
    09/05/14 | Annotating synapses in large EM datasets.
    Plaza SM, Parag T, Huang G, Olbris DJ, Saunders MA, Rivlin PK
    arXiv. 2014 Sep 5:arXiv:1409.1801 [q-bio.QM]

    Reconstructing neuronal circuits at the level of synapses is a central problem in neuroscience and becoming a focus of the emerging field of connectomics. To date, electron microscopy (EM) is the most proven technique for identifying and quantifying synaptic connections. As advances in EM make acquiring larger datasets possible, subsequent manual synapse identification ({\em i.e.}, proofreading) for deciphering a connectome becomes a major time bottleneck. Here we introduce a large-scale, high-throughput, and semi-automated methodology to efficiently identify synapses. We successfully applied our methodology to the Drosophila medulla optic lobe, annotating many more synapses than previous connectome efforts. Our approaches are extensible and will make the often complicated process of synapse identification accessible to a wider-community of potential proofreaders.

    View Publication Page
    09/05/14 | Automatic neuron type identification by neurite localization in the Drosophila medulla.
    Plaza SM, Zhao T
    arXiv. 2014 Sep 5:arXiv:1409.1892 [q-bio.NC]

    Mapping the connectivity of neurons in the brain (i.e., connectomics) is a challenging problem due to both the number of connections in even the smallest organisms and the nanometer resolution required to resolve them. Because of this, previous connectomes contain only hundreds of neurons, such as in the C.elegans connectome. Recent technological advances will unlock the mysteries of increasingly large connectomes (or partial connectomes). However, the value of these maps is limited by our ability to reason with this data and understand any underlying motifs. To aid connectome analysis, we introduce algorithms to cluster similarly-shaped neurons, where 3D neuronal shapes are represented as skeletons. In particular, we propose a novel location-sensitive clustering algorithm. We show clustering results on neurons reconstructed from the Drosophila medulla that show high-accuracy.

    View Publication Page
    09/05/14 | Identifying synapses using deep and wide multiscale recursive networks.
    Huang G, Plaza SM
    arXiv. 2014 Sep 5:arXiv:1409.1789 [cs.CV]

    In this work, we propose a learning framework for identifying synapses using a deep and wide multi-scale recursive (DAWMR) network, previously considered in image segmentation applications. We apply this approach on electron microscopy data from invertebrate fly brain tissue. By learning features directly from the data, we are able to achieve considerable improvements over existing techniques that rely on a small set of hand-designed features. We show that this system can reduce the amount of manual annotation required, in both acquisition of training data as well as verification of inferred detections.

    View Publication Page
    09/03/14 | Focused proofreading: efficiently extracting connectomes from segmented EM images.
    Plaza SM
    arXiv. 2014 Sep 3:arXiv:1409.1199 [q-bio.QM]

    Pixel and superpixel classifiers have become essential tools for EM segmentation algorithms. Training these classifiers remains a major bottleneck primarily due to the requirement of completely annotating the dataset which is tedious, error-prone and costly. In this paper, we propose an interactive learning scheme for the superpixel classifier for EM segmentation. Our algorithm is "active semi-supervised" because it requests the labels of a small number of examples from user and applies label propagation technique to generate these queries. Using only a small set (<20%) of all datapoints, the proposed algorithm consistently generates a classifier almost as accurate as that estimated from a complete groundtruth. We provide segmentation results on multiple datasets to show the strength of these classifiers.

    View Publication Page
    06/06/14 | Small sample learning of superpixel classifiers for EM segmentation- extended version.
    Parag T, Plaza SM, Scheffer LK
    arXiv. 2014 Jun 6:arXiv:1406.1774 [cs.CV]

    Pixel and superpixel classifiers have become essential tools for EM segmentation algorithms. Training these classifiers remains a major bottleneck primarily due to the requirement of completely annotating the dataset which is tedious, error-prone and costly. In this paper, we propose an interactive learning scheme for the superpixel classifier for EM segmentation. Our algorithm is "active semi-supervised" because it requests the labels of a small number of examples from user and applies label propagation technique to generate these queries. Using only a small set (<20%) of all datapoints, the proposed algorithm consistently generates a classifier almost as accurate as that estimated from a complete groundtruth. We provide segmentation results on multiple datasets to show the strength of these classifiers.

    View Publication Page
    06/05/14 | A context-aware delayed agglomeration framework for EM segmentation.
    Parag T, Chakraborty A, Plaza SM
    arXiv. 2014 Jun 5:arXiv:1406.1476 [cs.CV]

    This paper proposes a novel agglomerative framework for Electron Microscopy (EM) image (or volume) segmentation. For the overall segmentation methodology, we propose a context-aware algorithm that clusters the over-segmented regions of different sub-classes (representing different biological entities) in different stages. Furthermore, a delayed scheme for agglomerative clustering, which postpones the merge of newly formed bodies, is also proposed to generate a more confident boundary prediction. We report significant improvements in both segmentation accuracy and speed attained by the proposed approaches over existing standard methods on both 2D and 3D datasets.

    View Publication Page
    04/04/14 | Graph-based active learning of agglomeration (GALA): a Python library to segment 2D and 3D neuroimages
    Nunez-Iglesias J, Kennedy R, Plaza SM, Chakraborty A, William T. Katz
    Frontiers in Neuroinformatics. 2014 Apr 4;8:34. doi: 10.3389/fninf.2014.00034

    The aim in high-resolution connectomics is to reconstruct complete neuronal connectivity in a tissue. Currently, the only technology capable of resolving the smallest neuronal processes is electron microscopy (EM). Thus, a common approach to network reconstruction is to perform (error-prone) automatic segmentation of EM images, followed by manual proofreading by experts to fix errors. We have developed an algorithm and software library to not only improve the accuracy of the initial automatic segmentation, but also point out the image coordinates where it is likely to have made errors. Our software, called gala (graph-based active learning of agglomeration), improves the state of the art in agglomerative image segmentation. It is implemented in Python and makes extensive use of the scientific Python stack (numpy, scipy, networkx, scikit-learn, scikit-image, and others). We present here the software architecture of the gala library, and discuss several designs that we consider would be generally useful for other segmentation packages. We also discuss the current limitations of the gala library and how we intend to address them.

    View Publication Page