A1920 × 1080 25FPS, 2.4TOPS/W Unified Optical Flow and Depth 6D Vision Processor for Energy-Efficient, Low Power Autonomous Navigation.
Ziyun LiJingcheng WangDennis SylvesterDavid T. BlaauwHun-Seok KimPublished in: VLSI Circuits (2018)
Keyphrases
- low power
- energy efficient
- autonomous navigation
- high speed
- ego motion
- optical flow
- single chip
- computer vision
- wireless sensor networks
- low cost
- real time
- energy consumption
- energy efficiency
- power consumption
- sensor networks
- image sequences
- gate array
- field of view
- camera motion
- frame rate
- d scene
- vision system
- motion parameters
- flow field
- motion field
- stereo camera
- mobile robot
- base station
- motion estimation
- path planning
- motion model
- dynamic environments
- depth map
- depth information
- vector field
- motion segmentation
- sensor nodes
- moving camera
- optic flow
- moving objects
- image processing
- data transmission
- routing protocol
- dynamic scenes
- parallel processing
- data center
- video sequences
- three dimensional
- cmos technology
- sensor data