• Publications
  • Influence
Is BERT Really Robust? Natural Language Attack on Text Classification and Entailment
TLDR
We present the TextFooler, a general attack framework, to generate natural adversarial texts for text classification and textual entailment tasks. Expand
  • 72
  • 29
  • PDF
Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailment
TLDR
We present TextFooler, a simple but strong baseline to generate natural adversarial text that outperforms state-of-the-art attacks in terms of success rate and perturbation rate. Expand
  • 71
  • 18
Hybrid Heterogeneous Transfer Learning through Deep Learning
TLDR
We present a new transfer learning framework called Hybrid Heterogeneous Transfer Learning (HHTL), which allows the corresponding instances across domains to be biased in either the source or target domain. Expand
  • 117
  • 16
  • PDF
Heterogeneous Domain Adaptation for Multiple Classes
TLDR
In this paper, we present an efficient multi-class heterogeneous domain adaptation method, where data from source and target domains are represented by heterogeneous features of different dimensions. Expand
  • 100
  • 15
  • PDF
Exploit Bounding Box Annotations for Multi-Label Object Recognition
TLDR
In this paper, we incorporate local information to enhance the feature discriminative power and generalization ability of the multi-view multiinstance learning problem. Expand
  • 75
  • 10
  • PDF
MIML-FCN+: Multi-Instance Multi-Label Learning via Fully Convolutional Networks with Privileged Information
TLDR
We propose a two-stream fully convolutional network, named MIML-FCN+, unified by a novel PI loss to solve the problem of multi-instance multi-label learning with privileged bags. Expand
  • 33
  • 7
  • PDF
Does Deployment to Iraq and Afghanistan Affect Respiratory Health of US Military Personnel?
Objective: To evaluate the association between postdeployment respiratory conditions and deployment to Iraq or Afghanistan. Methods: We linked deployment history of US military personnel withExpand
  • 46
  • 5
A2J: Anchor-to-Joint Regression Network for 3D Articulated Pose Estimation From a Single Depth Image
TLDR
For 3D hand and body pose estimation task in depth image, a novel anchor-based approach termed Anchor-to-Joint regression network (A2J) with the end to-end learning ability is proposed. Expand
  • 27
  • 5
  • PDF
Structured AutoEncoders for Subspace Clustering
TLDR
This paper proposed a novel neural network for subspace clustering by introducing a new deep model—Structured AutoEncoder (StructAE). Expand
  • 155
  • 4
  • PDF
Feature Disentangling Machine - A Novel Approach of Feature Selection and Disentangling in Facial Expression Analysis
TLDR
We propose a unified MTL framework, named Feature Disentangling Machine (FDM), to simultaneously select and disentangle common features and expression-specific features. Expand
  • 51
  • 4
  • PDF