W
Weidi Xie
Researcher at University of Oxford
Publications - 107
Citations - 6398
Weidi Xie is an academic researcher from University of Oxford. The author has contributed to research in topics: Computer science & Segmentation. The author has an hindex of 22, co-authored 65 publications receiving 3588 citations.
Papers
More filters
Proceedings ArticleDOI
Open-vocabulary Semantic Segmentation with Frozen Vision-Language Models
TL;DR: This paper introduces Fusioner, with a lightweight, transformer-based fusion module, that pairs the frozen visual representation with language concept through a handful of image segmentation data, and demonstrates superior performance over previous models.
Posted Content
Feature Tracking Cardiac Magnetic Resonance via Deep Learning and Spline Optimization
Davis M. Vigneault,Davis M. Vigneault,Davis M. Vigneault,Weidi Xie,David A. Bluemke,J. Alison Noble +5 more
TL;DR: In this article, the authors proposed a fully automated technique whereby a CMR image sequence is first segmented with a deep, fully convolutional neural network (CNN) architecture, and quadratic basis splines are fitted simultaneously across all cardiac frames using least squares optimization.
Posted Content
{\Omega}-Net (Omega-Net): Fully Automatic, Multi-View Cardiac MR Detection, Orientation, and Segmentation with Deep Neural Networks
TL;DR: A novel convolutional neural network architecture for simultaneous detection, transformation into a canonical orientation, and semantic segmentation is presented, believed to represent a substantive advancement over prior approaches, with implications for biomedical image segmentation more generally.
Posted Content
Class-Agnostic Counting
TL;DR: In this paper, a generic matching network (GMN) is proposed to count any object in a class-agnostic manner, and the model achieves competitive performance on cell and crowd counting datasets.
Proceedings ArticleDOI
Segmenting Moving Objects via an Object-Centric Layered Representation
Jun Xie,Weidi Xie,A. Zisserman +2 more
TL;DR: An object-centric segmentation model with a depth-ordered layer representation implemented using a variant of the transformer architecture that ingests optical layers, which can effectively discover multiple moving objects and handle mutual occlusions.