Deep Learning Approach for Pose Estimation Talk #23444 MSc Kanter - - PowerPoint PPT Presentation

deep learning approach for pose estimation
SMART_READER_LITE
LIVE PREVIEW

Deep Learning Approach for Pose Estimation Talk #23444 MSc Kanter - - PowerPoint PPT Presentation

Deep Learning Approach for Pose Estimation Talk #23444 MSc Kanter van Deurzen Introduction Delft Robotics Kanter van Deurzen Founded 2014 CTO Delft Robotics Staff of 11 Specialized in Vision guided Robotics through AI Winner Amazon


slide-1
SLIDE 1

Deep Learning Approach for Pose Estimation

Talk #23444 MSc Kanter van Deurzen

slide-2
SLIDE 2

Introduction

Delft Robotics Founded 2014 Staff of 11 Specialized in Vision guided Robotics through AI Winner Amazon Picking Challenge 2016 Kanter van Deurzen CTO Delft Robotics

slide-3
SLIDE 3

Application

Bin-picking

  • One type of object
  • Multiple objects
  • Restricted area

Order-picking

  • Large number of different objects
  • Multiple objects
  • Restricted area

Service robots

  • Large number of different objects
  • No area restriction

Any orientation possible (6DOF)!!

slide-4
SLIDE 4

Typical pipeline

Data acquisition

  • 2D
  • 3D

Segmentation

  • Deep learning

Rough pose estimation

  • Ransac
  • 4PCS
  • Fast global

registration Pose

  • ptimization
  • ICP

Grasp pose estimation

slide-5
SLIDE 5

CAD pipeline

  • 1. Locate object in 2D/3D
  • 2. Use CAD model to find global optimum of object

pose

  • 3. Use CAD model to refine pose locally
  • 4. Determine grasp pose using estimated pose
slide-6
SLIDE 6

CAD pipeline

  • Advantages:

– Complete 6DOF pose is known – Pose in gripper approximately known – Scales to other objects

  • Disadvantages:

– Symmetry causes ambiguity, risk of local minima – Requires CAD model

slide-7
SLIDE 7

Non-rigid/CAD-less pipeline

  • 1. Locate object in 2D/3D
  • 2. Fit shape of gripper anywhere on object
slide-8
SLIDE 8

Non-rigid/CAD-less pipeline

  • Advantages:

– No CAD model necessary – No knowledge of object necessary

  • Disadvantages:

– Pose in gripper unknown – No knowledge of object used (fragility/weight/...)

slide-9
SLIDE 9

Why deep learning?

  • Rough pose estimation is slow
  • Rough pose estimation risks local minima
slide-10
SLIDE 10

Attempted DL Approaches

  • Classification
  • Regression
  • Key-points
slide-11
SLIDE 11

Classification

  • Concept

– Classify pose as one of X classes

Class 1 Class 2 Class 3 Class ...

slide-12
SLIDE 12

Classification results

  • Conditions:

– Rendered images – 3 axes rotations – 10 classes per axis (36 degrees per class) – For only 1 axis, great (Z-axis, 99% accuracy)

  • Conclusions:

– Scales badly to multiple axes (10 classes per axis, 10*10*10 = 1000 total classes).

slide-13
SLIDE 13

Regression

  • Treat as a regression problem
  • Regress 4 values (queternion)
  • Use RGB(D) as input

Ground truth Predicted pose

slide-14
SLIDE 14

Regression results

  • Conclusions:

– Works well on asymmetric objects – Symmetric objects cause ambiguity – Difficult to find correct loss function

slide-15
SLIDE 15

Key-points

  • Concept:

– Recognize 3 or more keypoints – Determine pose based on these keypoints – Refine using (local) heuristics

slide-16
SLIDE 16

Key-points results

  • Conclusions:

– Easily integrated end-to-end in network – Works only for well-defined keypoints, which is non-trivial – Results in good initial guesses

slide-17
SLIDE 17

Future research

  • Investigate other DL approaches (
  • Process point clouds directly with DL

– challenges: data, data, data, lack of research

  • Use other forms of DL (reinforcement learning? GANs?)
slide-18
SLIDE 18

Thank you for your attention

Kantervan Deurzen k.vandeurzen@delftrobotics.com 0651753705