Deep neural networks with attention mechanism for monocular depth estimation on embedded devices.
Siping LiuXiaohan TuCheng XuRenfa LiPublished in: Future Gener. Comput. Syst. (2022)
Keyphrases
- embedded devices
- depth estimation
- attention mechanism
- depth map
- stereo vision
- depth estimates
- embedded systems
- visual attention
- stereo matching
- scene understanding
- image sequences
- dynamic scenes
- stereo pair
- depth information
- saliency map
- mobile devices
- real scenes
- multi view
- disparity map
- pose estimation
- super resolution
- feature matching
- optical flow
- specular reflection
- d scene
- low cost
- computer vision
- data storage
- limited memory
- feature vectors
- flash memory
- computational complexity
- high quality
- three dimensional