• Publications
  • Influence
Parallel Separable 3D Convolution for Video and Volumetric Data Understanding
TLDR
This paper mathematically justifies the need of parallel streams (Pm) to replace a single 3D convolution layer through tensor decomposition and empirically shows that PmSCn is applicable to different backbone architectures, such as ResNet, DenseNet, and UNet, for different applications.
ICON: An interactive approach to train deep neural networks for segmentation of neuronal structures
TLDR
The experiments show that an interactively-trained pixel classifier produces better region segmentation results on Electron Microscopy (EM) images than those generated by a network of the same architecture trained offline on exhaustive ground-truth labels.
Scalable Interactive Visualization for Connectomics
TLDR
This work presents the BUTTERFLY middleware, a scalable platform that can handle massive data for interactive visualization in connectomics and demonstrates scalability and extendability with a series of open source Web-based applications for every step of the typical connectomics workflow.
Consistent Recurrent Neural Networks For 3d Neuron Segmentation.
TLDR
A recurrent network for 3D reconstruction of neurons that sequentially generates binary masks for every object in an image with spatio-temporal consistency is presented and achieved state-of-the-art performance on the SNEMI3D challenge.
Vice
TLDR
This paper presents the design and implementation of an analytics framework that streamlines proofreading, focusing on connectivity-related errors, with automated likely-error detection and synapse clustering that drives the proofreading effort with highly interactive 3D visualizations.
SPARTA : Fast global planning of collision-avoiding robot trajectories
TLDR
This work presents an algorithm for obtaining collision-avoiding robot trajectories called SPARTA (SPline-based ADMM for Robot Trajectory Avoidance), using a recently developed generalization of ADMM, the Three Weight Algorithm, which can be elegantly understood as a message passing algorithm.
An Interactive Deep Learning Toolkit for Automatic Segmentation of Images
Image Segmentation, Neuron Segmentation, Interactive Segmentation, Deep Neural Networks, Multilayer Perceptron, Deep Learning
VICE: Visual Identification and Correction of Neural Circuit Errors
TLDR
This paper presents the design and implementation of an analytics framework that streamlines proofreading, focusing on connectivity‐related errors, with automated likely‐error detection and synapse clustering that drives the proofreading effort with highly interactive 3D visualizations.
Vice
TLDR
This paper presents the design and implementation of an analytics framework that streamlines proofreading, focusing on connectivity-related errors, with automated likely-error detection and synapse clustering that drives the proofreading effort with highly interactive 3D visualizations.