End-to-End Tracking and Semantic Segmentation Using Recurrent Neural Networks.
Peter OndruskaJulie DequaireDominic Zeng WangIngmar PosnerPublished in: CoRR (2016)
Keyphrases
- end to end
- recurrent neural networks
- semantic segmentation
- street scenes
- neural network
- conditional random fields
- superpixels
- weakly supervised
- scene classification
- feed forward
- real time
- echo state networks
- artificial neural networks
- object categories
- image understanding
- particle filter
- visual tracking
- object tracking
- image sequences
- object classes
- pascal voc
- natural images
- object detection
- multiscale
- multiple objects
- bounding box
- mean shift
- topic models