• Publications
  • Influence
Adversarial Examples: Attacks and Defenses for Deep Learning
TLDR
We review recent findings on adversarial examples for DNNs, summarize the methods for generating adversarial example, and propose a taxonomy of these methods. Expand
  • 570
  • 29
  • PDF
Single Shot Text Detector with Regional Attention
TLDR
We present a novel single-shot text detector that directly outputs word-level bounding boxes in a natural image. Expand
  • 176
  • 15
  • PDF
GRAM-CNN: a deep learning approach with local context for named entity recognition in biomedical text
TLDR
We propose a novel end-to-end deep learning approach for biomedical NER tasks that leverages the local contexts based on n-gram character and word embeddings via Convolutional Neural Network (CNN). Expand
  • 57
  • 3
Statistical learning for semantic parsing: A survey
TLDR
A long-term goal of Artificial Intelligence (AI) is to provide machines with the capability of understanding natural language. Expand
  • 4
  • 2
GraphBTM: Graph Enhanced Autoencoded Variational Inference for Biterm Topic Model
TLDR
We propose a novel way called GraphBTM to represent bitermms as graphs and design a Graph Convolutional Networks (GCNs) with residual connections to extract transitive features from biterms. Expand
  • 11
  • 1
  • PDF
A Batch Normalized Inference Network Keeps the KL Vanishing Away
TLDR
We propose to let the KL follow a distribution across the whole dataset, and analyze that it is sufficient to prevent posterior collapse by keeping the expectation of the KL's distribution positive. Expand
  • 4
  • 1
  • PDF
Improving Question Generation with Sentence-level Semantic Matching and Answer Position Inferring
TLDR
We propose a neural question generation model with two general modules: sentence-level semantic matching and answer position inferring. Expand
  • 4
  • 1
  • PDF
Local Contextual Attention with Hierarchical Structure for Dialogue Act Recognition
TLDR
We propose a hierarchical model based on self-attention to capture intra-sentence and contextual semantic information by incorporating the relative position information between utterances. Expand
  • 2
  • 1
  • PDF
Character Sequence-to-Sequence Model with Global Attention for Universal Morphological Reinflection
TLDR
This paper presents a neural network based approach for the CoNLLSIGMORPHON-2017 Shared Task 1 on morphological reinflection and achieved good results in some languages. Expand
  • 2
  • PDF
Learning Fast and Slow: PROPEDEUTICA for Real-time Malware Detection
TLDR
We introduce and evaluate PROPEDEUTICA, a novel methodology and framework for efficient and effective real-time malware detection, leveraging the best of conventional machine learning (ML) and deep learning (DL) algorithms. Expand
  • 4
  • PDF