Toward Affordance Detection and Ranking on Novel Objects for Real-World Robotic Manipulation.
Fu-Jen ChuRuinian XuLandan SeguinPatricio A. VelaPublished in: IEEE Robotics Autom. Lett. (2019)
Keyphrases
- real world
- heterogeneous information networks
- wide range
- complex background
- bounding box
- physical objects
- real time
- object detection
- detection algorithm
- synthetic data
- matched filters
- man made
- false alarms
- object classes
- data mining
- object categories
- partial occlusion
- manipulation tasks
- cluttered scenes
- automatic detection
- detection accuracy
- object detectors
- information networks
- complex scenes
- target object
- ranking functions
- robotic systems
- detection method
- data objects
- object models
- information retrieval systems
- multiple types
- mobile robot
- object model
- individual objects
- video sequences
- search engine
- learning to rank