• Publications
  • Influence
Saliency detection by multi-context deep learning
TLDR
We propose a multi-context deep learning framework to model saliency of objects in images by taking global and local context into account. Expand
  • 664
  • 94
  • PDF
Learning Deep Feature Representations with Domain Guided Dropout for Person Re-identification
TLDR
Learning generic and robust feature representations with data from multiple domains for the same problem is of great value, especially for the problems that have multiple datasets but none of them are large enough. Expand
  • 751
  • 88
  • PDF
StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks
TLDR
We propose Stacked Generative Adversarial Networks (StackGANs) aimed at generating high-resolution photo-realistic images. Expand
  • 382
  • 75
  • PDF
Orientation Invariant Feature Embedding and Spatial Temporal Regularization for Vehicle Re-identification
TLDR
In this paper, we tackle the vehicle Re-identification (ReID) problem which is of great importance in urban surveillance and can be used for multiple applications. Expand
  • 160
  • 41
  • PDF
Learning Spatial Regularization with Image-Level Supervisions for Multi-label Image Classification
TLDR
We propose a unified deep neural network for multi-label image classification, which exploits both semantic and spatial relations between labels with only image-level supervisions. Expand
  • 143
  • 32
  • PDF
Learning Feature Pyramids for Human Pose Estimation
TLDR
We design a Pyramid Residual Module (PRMs) to enhance the invariance in scales of DCNNs. Expand
  • 259
  • 31
  • PDF
T-CNN: Tubelets With Convolutional Neural Networks for Object Detection From Videos
TLDR
We propose a deep learning framework that incorporates temporal and contextual information from tubelets obtained in videos, which dramatically improves the baseline performance of existing still-image detection frameworks when they are applied to videos. Expand
  • 206
  • 20
  • PDF
Object Detection from Video Tubelets with Convolutional Neural Networks
TLDR
In this work, we introduce a complete framework for the VID task based on still-image object detection and general object tracking. Expand
  • 238
  • 17
  • PDF
Understanding pedestrian behaviors from stationary crowd groups
TLDR
In this paper, a novel model is proposed for pedestrian behavior modeling by including stationary crowd groups as a key component. Expand
  • 147
  • 17
  • PDF
Online Multi-object Tracking Using CNN-Based Single Object Tracker with Spatial-Temporal Attention Mechanism
TLDR
We introduce spatial-temporal attention mechanism (STAM) to handle the drift caused by occlusion and interaction among targets. Expand
  • 161
  • 15
  • PDF