• Publications
  • Influence
From captions to visual concepts and back
TLDR
This paper presents a novel approach for automatically generating image descriptions: visual detectors, language models, and multimodal similarity models learnt directly from a dataset of image captions. Expand
  • 990
  • 85
  • PDF
A Neural Network Approach to Context-Sensitive Generation of Conversational Responses
TLDR
We present a novel response generation system that can be trained end to end on large quantities of unstructured Twitter conversations. Expand
  • 695
  • 69
  • PDF
Mitigating Unwanted Biases with Adversarial Learning
TLDR
We present a framework for mitigating such biases by including a variable for the group of interest and simultaneously learning a predictor and an adversary. Expand
  • 306
  • 42
  • PDF
Spoken Language Derived Measures for Detecting Mild Cognitive Impairment
TLDR
We present results on the utility of such markers in discriminating between healthy elderly subjects and subjects with mild cognitive impairment (MCI). Expand
  • 216
  • 32
  • PDF
Visual Storytelling
TLDR
We introduce the first dataset for sequential vision-to-language, and explore how this data may be used for the task of visual storytelling. Expand
  • 202
  • 30
  • PDF
Open Domain Targeted Sentiment
TLDR
We propose a novel approach to sentiment analysis for a low resource setting, using only a sentiment lexicon as an external resource. Expand
  • 128
  • 23
  • PDF
Language Models for Image Captioning: The Quirks and What Works
TLDR
We compare the merits of these different language modeling approaches for the first time by using the same state-ofthe-art CNN as input. Expand
  • 216
  • 21
  • PDF
VQA: Visual Question Answering
TLDR
We propose the task of free-form and open-ended Visual Question Answering (VQA). Expand
  • 180
  • 20
Midge: Generating Image Descriptions From Computer Vision Detections
TLDR
This paper introduces a novel generation system that composes humanlike descriptions of images from computer vision detections. Expand
  • 345
  • 19
  • PDF
Exploring Nearest Neighbor Approaches for Image Captioning
TLDR
We explore a variety of nearest neighbor approaches for image captioning. Expand
  • 153
  • 19
  • PDF
...
1
2
3
4
5
...