Real-Time Semantic Segmentation with Label Propagation

  title={Real-Time Semantic Segmentation with Label Propagation},
  author={Rasha Sheikh and Martin Garbade and Juergen Gall},
  booktitle={ECCV Workshops},
Despite of the success of convolutional neural networks for semantic image segmentation, CNNs cannot be used for many applications due to limited computational resources. Even efficient approaches based on random forests are not efficient enough for real-time performance in some cases. In this work, we propose an approach based on superpixels and label propagation that reduces the runtime of a random forest approach by factor 192 while increasing the segmentation accuracy. 
Semi-supervised Semantic Mapping Through Label Propagation with Semantic Texture Meshes
This work proposes an efficient method of establishing consensus between individual segmentations by iteratively retraining semantic segmentation with the information stored within the map and using the retrained segmentation to re-fuse the semantics.
Uncertainty Modeling of Contextual-Connections Between Tracklets for Unconstrained Video-Based Face Recognition
The Uncertainty-Gated Graph (UGG), which conducts graph-based identity propagation between tracklets, which are represented by nodes in a graph, is proposed, which explicitly models the uncertainty of the contextual connections by adaptively updating the weights of the edge gates according to the identity distributions of the nodes during inference.
Person Search in Videos with One Portrait Through Visual and Temporal Links
A novel framework is proposed, which takes into account the identity invariance along a tracklet, thus allowing person identities to be propagated via both the visual and the temporal links and remarkably outperforms mainstream person re-id methods.


Fast label propagation for real-time superpixels for video content
This paper proposes a fast propagation scheme that is based on sparse feature tracking and mesh-based image warping that speeds up the propagation process by a factor of 100 while producing a comparable segmentation quality.
Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
This work brings together methods from DCNNs and probabilistic graphical models for addressing the task of pixel-level classification by combining the responses at the final DCNN layer with a fully connected Conditional Random Field (CRF).
CURFIL: Random Forests for Image Labeling on GPU
This work presents an open-source implementation that significantly accelerates both random forest learning and prediction for image labeling of RGB-D and RGB images on GPU when compared to an optimized multi-core CPU implementation.
Learning Hierarchical Features for Scene Labeling
A method that uses a multiscale convolutional network trained from raw pixels to extract dense feature vectors that encode regions of multiple sizes centered on each pixel, alleviates the need for engineered features, and produces a powerful representation that captures texture, shape, and contextual information.
Combining Appearance and Structure from Motion Features for Road Scene Understanding
A framework for pixel-wise object segmentation of road scenes that combines motion and appearance features that is designed to handle street-level imagery such as that on Google Street View and Microsoft Bing Maps is presented.
SegNet: A Deep Convolutional Encoder-Decoder Architecture for Robust Semantic Pixel-Wise Labelling
The results show that SegNet achieves state-of-the-art performance even without use of additional cues such as depth, video frames or post-processing with CRF models.
Semantic Segmentation of Urban Scenes Using Dense Depth Maps
The result shows that only using dense depth information, this framework for semantic scene parsing and object recognition based on dense depth maps can achieve overall better accurate segmentation and recognition than that from sparse 3D features or appearance, advancing state-of-the-art performance.
Very Deep Convolutional Networks for Large-Scale Image Recognition
This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.
Local Label Descriptor for Example Based Semantic Image Labeling
The concept of local label descriptor is introduced, which is a concatenation of label histograms for each cell in a patch, which alleviates the label patch misalignment issue in combining structured label predictions for semantic image labeling.
Label propagation in complex video sequences using semi-supervised learning
A novel directed graphical model for label propagation in lengthy and complex video sequences using a hybrid of generative propagation and discriminative classification in a pseudo time-symmetric video model to achieve a conservative labelling of the video.