Chapter 8 Robotic Robotic 1 Robotics is the
Chapter 8 Robotic
Robotic 1) Robotics is the intelligent connection of perception action. 2) A robotic is anything that is surprisingly (moving target) animate. 3) perceptual (S/W) + motor task (H/W) [action] • operate in the real world : searching and backtracking can be costly • we need operating in a simulate world with full information for an optimal plan by best-first search • we can checked preconditions of the operators using perception to perform action • real time search : p. 562 A* algorithm, RTA* (Korf 1988) Artificial Intelligence 2 Chapter 7
Artificial Intelligence 3 Chapter 7
Vision 2 D 3 D signal processing : enhance the image measurement analysis : for image containing a single object determining the 2 D extent of the object depicted pattern recognition : for single object images, classify the object into category image understanding : for image containing many objects in the image, classify them, build 3 D model of the scene. see Figure 14. 8 p. 367 Artificial Intelligence 4 Chapter 7
Artificial Intelligence 5 Chapter 7
Vision • Problem : • ambiguous image : see Figure 21. 2 p. 564 • Figure 21. 3 p. 565 using low level knowledge to interpret an image • image factor, sensor fusion : color, reflectance, shading • Figure 21. 4 p. 565 using high level knowledge to interpret an image (a) use surroundings objects to help (b) baseball, log in a fireplace, amoeba, [egg, bacon, and plate] • Figure 21. 5 p. 567 Image understanding • analog signal Image 2 D features 3 D features • 3 D composite objects Object identification Artificial Intelligence 6 Chapter 7
Artificial Intelligence 7 Chapter 7
Artificial Intelligence 8 Chapter 7
Artificial Intelligence 9 Chapter 7
Artificial Intelligence 10 Chapter 7
Speech Recognition • speaker dependence (we can train the system) / speaker independence • continuous / isolated word speech • real time SPHINX (1988) / offline processing • large (difficult) / small vocabulary • broad (difficult) / narrow grammar: TANGORA (1985) 20000 words vocabulary • HMM (Hidden Markov Modeling) SPHINX system – – statistical learning method HMM is a collection of states and transitions Artificial Intelligence 11 Chapter 7
Speech Recognition HMM (Hidden Markov Modeling) SPHINX system statistical learning method HMM is a collection of states and transitions the problem of decoding a speech waveform turns into the problem of finding the most likely path (set of transitions) through an appropriate KMM. Artificial Intelligence 12 Chapter 7
Action • • • p. 569 : navigation around the world planning routes / path planning reaching desired destinations without bumping into things see Figure 21. 6– 21. 9 p. 570 -571 constructing a visibility graph • configuration space / C-space (Lozano-Perez 1984) – basic idea is to reduce the robot to a point P and do path planning in an artificially constructed space – rotation (X, Y, ) • obstacles can be transformed into 3 D C-space objects, visibility graph can be created and searched. Artificial Intelligence 13 Chapter 7
Artificial Intelligence 14 Chapter 7
Artificial Intelligence 15 Chapter 7
Artificial Intelligence 16 Chapter 7
Manipulation • end-effectors (two-gripper) / a human like hand • pick-and-place : grasp and object and move it to a specific location see Figure 21. 10 -21. 11 p. 572 -573 • Figure 21. 11 -21. 12 (a) naive strategy for grasping and placement • Figure 21. 11 -21. 12 (b) clever strategy for grasping and placement • planning p. 332 e. g. Block world ON(A, B) HOLDING , ARMEMPTY Artificial Intelligence 17 Chapter 7
Artificial Intelligence 18 Chapter 7
Artificial Intelligence 19 Chapter 7
Manipulation • planning p. 332 e. g. Block world ON(A, B) HOLDING , ARMEMPTY • Components of a planning system 1) 2) 3) 4) 5) choose the best rule to applying rules see Figure 13. 2 -13. 3 p. 336 -337 detecting a solution detecting dead ends repairing an almost correct solution Artificial Intelligence 20 Chapter 7
Artificial Intelligence 21 Chapter 7
Artificial Intelligence 22 Chapter 7
The End Artificial Intelligence 23 Chapter 7
- Slides: 23
