Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
Institut national de recherche en informatique et en automatique · CentraleSupélec · +2 more institutions
Abstract
Deep Convolutional Neural Networks (DCNNs) have recently shown state of the art performance in high level vision tasks, such as image classification and object detection. This work brings together methods from DCNNs and probabilistic graphical models for addressing the task of pixel-level classification (also called "semantic image segmentation"). We show that responses at the final layer of DCNNs are not sufficiently localized for accurate object segmentation. This is due to the very invariance properties that make DCNNs good for high level tasks. We overcome this poor localization property of deep networks by combining the responses at the final DCNN layer with a fully connected Conditional Random Field…
Citation impact
- FWCI
- —
- Percentile
- —
- References
- 64
Authors
5- LCLiang-Chieh ChenCorresponding
Institut national de recherche en informatique et en automatique, CentraleSupélec, Google (United States), University of California, Los Angeles
- GPGeorge Papandreou
Google (United States), University of California, Los Angeles, CentraleSupélec, Institut national de recherche en informatique et en automatique
- IKIasonas Kokkinos
University of California, Los Angeles, Institut national de recherche en informatique et en automatique, CentraleSupélec, Google (United States)
- KMKevin Murphy
Institut national de recherche en informatique et en automatique, University of California, Los Angeles, CentraleSupélec, Google (United States)
- AYAlan Yuille
Institut national de recherche en informatique et en automatique, University of California, Los Angeles, CentraleSupélec, Google (United States)
Topics & keywords
- Conditional random field
- Computer science
- Artificial intelligence
- Convolutional neural network
- CRFS
- Segmentation
- Pattern recognition (psychology)
- Image segmentation
- No poverty