Traffic scene semantic segmentation using self-attention mechanism and bi-directional GRU to correlate context.
Min YanJunzheng WangJing LiKe ZhangZimu YangPublished in: Neurocomputing (2020)
Keyphrases
- bi directional
- semantic segmentation
- scene classification
- attention mechanism
- superpixels
- three dimensional
- conditional random fields
- image sequences
- object recognition
- input image
- image set
- visual attention
- weakly supervised
- natural scenes
- associative memory
- machine learning
- biologically inspired
- video sequences
- saliency map
- single image
- human computer interaction
- image classification
- information extraction
- image segmentation