Timezone: »
Deep structured output learning shows great promise in tasks like semantic image segmentation. We proffer a new, efficient deep structured model learning scheme, in which we show how deep Convolutional Neural Networks (CNNs) can be used to directly estimate the messages in message passing inference for structured prediction with Conditional Random Fields CRFs). With such CNN message estimators, we obviate the need to learn or evaluate potential functions for message calculation. This confers significant efficiency for learning, since otherwise when performing structured learning for a CRF with CNN potentials it is necessary to undertake expensive inference for every stochastic gradient iteration. The network output dimension of message estimators is the same as the number of classes, rather than exponentially growing in the order of the potentials. Hence it is more scalable for cases that a large number of classes are involved. We apply our method to semantic image segmentation and achieve impressive performance, which demonstrates the effectiveness and usefulness of our CNN message learning method.
Author Information
Guosheng Lin (The University of Adelaide)
Chunhua Shen (University of Adelaide)
Ian Reid (University of Adelaide)
Anton van den Hengel (University of Adelaide)
More from the Same Authors
-
2020 Poster: Counterfactual Vision-and-Language Navigation: Unravelling the Unseen »
Amin Parvaneh · Ehsan Abbasnejad · Damien Teney · Javen Qinfeng Shi · Anton van den Hengel -
2020 Spotlight: Counterfactual Vision-and-Language Navigation: Unravelling the Unseen »
Amin Parvaneh · Ehsan Abbasnejad · Damien Teney · Javen Qinfeng Shi · Anton van den Hengel -
2020 Poster: SOLOv2: Dynamic and Fast Instance Segmentation »
Xinlong Wang · Rufeng Zhang · Tao Kong · Lei Li · Chunhua Shen -
2020 Poster: On the Value of Out-of-Distribution Testing: An Example of Goodhart's Law »
Damien Teney · Ehsan Abbasnejad · Kushal Kafle · Robik Shrestha · Christopher Kanan · Anton van den Hengel -
2019 Poster: Social-BiGAT: Multimodal Trajectory Forecasting using Bicycle-GAN and Graph Attention Networks »
Vineet Kosaraju · Amir Sadeghian · Roberto Martín-Martín · Ian Reid · Hamid Rezatofighi · Silvio Savarese -
2019 Poster: Unsupervised Scale-consistent Depth and Ego-motion Learning from Monocular Video »
Jiawang Bian · Zhichao Li · Naiyan Wang · Huangying Zhan · Chunhua Shen · Ming-Ming Cheng · Ian Reid -
2019 Poster: Multi-marginal Wasserstein GAN »
Jiezhang Cao · Langyuan Mo · Yifan Zhang · Kui Jia · Chunhua Shen · Mingkui Tan -
2017 Poster: Deep Subspace Clustering Networks »
Pan Ji · Tong Zhang · Hongdong Li · Mathieu Salzmann · Ian Reid -
2017 Poster: A Bayesian Data Augmentation Approach for Learning Deep Models »
Toan Tran · Trung Pham · Gustavo Carneiro · Lyle Palmer · Ian Reid -
2016 Poster: Image Restoration Using Very Deep Convolutional Encoder-Decoder Networks with Symmetric Skip Connections »
Xiaojiao Mao · Chunhua Shen · Yu-Bin Yang -
2014 Poster: Encoding High Dimensional Local Features by Sparse Coding Based Fisher Vectors »
Lingqiao Liu · Chunhua Shen · Lei Wang · Anton van den Hengel · Chao Wang -
2009 Poster: Positive Semidefinite Metric Learning with Boosting »
Chunhua Shen · Junae Kim · Lei Wang · Anton van den Hengel