• Publications
  • Influence
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation
TLDR
We show how to improve semantic segmentation through the use of contextual information, specifically, we explore ' patch-patch' context between image regions, and 'patch-background' context. Expand
  • 688
  • 55
  • PDF
Online Multi-Target Tracking Using Recurrent Neural Networks
TLDR
We present a novel approach to online multi-target tracking based on recurrent neural networks (RNNs). Tracking multiple objects in real-world scenes involves many challenges, including a) an a-priori unknown and time-varying number of targets, b) a continuous state estimation of all present targets, and a discrete combinatorial problem of data association. Expand
  • 318
  • 34
  • PDF
AffordanceNet: An End-to-End Deep Learning Approach for Object Affordance Detection
TLDR
We propose AffordanceNet, an end-to-end deep learning framework that can simultaneously detect multiple objects and their affordances from RGB images. Expand
  • 106
  • 18
  • PDF
Smart Mining for Deep Metric Learning
TLDR
In this paper, we propose a novel deep metric learning method that combines the triplet model and the global structure of the embedding space. Expand
  • 160
  • 12
  • PDF
Visual Odometry Revisited: What Should Be Learnt?
TLDR
In this work we present a monocular visual odometry (VO) algorithm which leverages geometry-based methods and deep learning. Expand
  • 17
  • 2
  • PDF
Bayesian Semantic Instance Segmentation in Open Set World
TLDR
A novel open-set semantic instance segmentation approach capable of segmenting all known and unknown object classes in images, based on the output of an object detector trained on known object classes. Expand
  • 17
  • PDF
Probabilistic tracking and recognition of nonrigid hand motion
  • H. Fei, Ian Reid
  • Computer Science
  • IEEE International SOI Conference. Proceedings…
  • 17 October 2003
TLDR
We divide and conquer tracking of articulated hand motion by decomposing complex motion into nonrigid motion and rigid motion. Expand
  • 18
  • PDF
Computer Vision -- ACCV 2014
TLDR
The five-volume set LNCS 9003--9007 constitutes the thoroughly refereed post-conference proceedings of the 12th Asian Conference on Computer Vision, ACCV 2014, held in Singapore, Singapore, in November 2014. Expand
  • 5
Structured Binary Neural Networks for Image Recognition
TLDR
We propose methods to train convolutional neural networks (CNNs) with both binarized weights and activations, leading to quantized models that are specifically friendly to mobile devices with limited power capacity and computation resources. Expand
  • 3
  • PDF
SPRINT: Subgraph Place Recognition for INtelligent Transportation
TLDR
We propose SPRINT, a scalable subgraph based HMM inference framework for large scale place recognition over databases containing millions of images. Expand
  • 1