DenseCap: Fully Convolutional Localization Networks for Dense Captioning

We introduce the dense captioning task, which requires a computer vision system to both localize and describe salient regions in images in natural language. The dense captioning task generalizes object detection when the descriptions consist of a single word, and Image Captioning when one predicted...

Full description

Saved in:
Bibliographic Details
Published in2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) pp. 4565 - 4574
Main Authors Johnson, Justin, Karpathy, Andrej, Li Fei-Fei
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2016
Subjects
Interpolation
Object detection
Predictive models
Proposals
Shape
Tensile stress
Visualization
Online AccessGet full text

Cover

More Information
Summary:We introduce the dense captioning task, which requires a computer vision system to both localize and describe salient regions in images in natural language. The dense captioning task generalizes object detection when the descriptions consist of a single word, and Image Captioning when one predicted region covers the full image. To address the localization and description task jointly we propose a Fully Convolutional Localization Network (FCLN) architecture that processes an image with a single, efficient forward pass, requires no external regions proposals, and can be trained end-to-end with a single round of optimization. The architecture is composed of a Convolutional Network, a novel dense localization layer, and Recurrent Neural Network language model that generates the label sequences. We evaluate our network on the Visual Genome dataset, which comprises 94,000 images and 4,100,000 region-grounded captions. We observe both speed and accuracy improvements over baselines based on current state of the art approaches in both generation and retrieval settings.
ISSN:1063-6919
DOI:10.1109/CVPR.2016.494