6-PACK: Category-level 6D Pose Tracker with Anchor-Based Keypoints (2019-10-23T00:00:00.000000Z)

TL;DR

6-PACK learns to compactly represent an object by a handful of 3D keypoints, based on which the interframe motion of an object instance can be estimated through keypoint matching, and substantially outperforms existing methods on the NOCS category-level 6D pose estimation benchmark.

Abstract

We present 6-PACK, a deep learning approach to category-level 6D object pose tracking on RGB-D data. Our method tracks in real time novel object instances of known object categories such as bowls, laptops, and mugs. 6-PACK learns to compactly represent an object by a handful of 3D keypoints, based on which the interframe motion of an object instance can be estimated through keypoint matching. These keypoints are learned end-to-end without manual supervision in order to be most effective for tracking. Our experiments show that our method substantially outperforms existing methods on the NOCS category-level 6D pose estimation benchmark and supports a physical robot to perform simple vision-based closed-loop manipulation tasks. Our code and video are available at https://sites.google.com/view/6packtracking.

Authors

S. Savarese

31 papers

Li Fei-Fei

37 papers

Yuke Zhu

15 papers

TL;DR

Abstract

Authors

References54 items

PoseRBPF: A Rao–Blackwellized Particle Filter for 6-D Object Pose Tracking

CornerNet-Lite: Efficient Keypoint based Object Detection

kPAM: KeyPoint Affordances for Category-Level Robotic Manipulation

DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion

Normalized Object Coordinate Space for Category-Level 6D Object Pose and Size Estimation

Segmentation-Driven 6D Object Pose Estimation

Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects

SilhoNet: An RGB Method for 6D Object Pose Estimation

SilhoNet: An RGB Method for 3D Object Pose Estimation and Grasp Planning.

Discovery of Latent 3D Keypoints via End-to-end Geometric Reasoning

DeepIM: Deep Iterative Matching for 6D Pose Estimation

DensePose: Dense Human Pose Estimation in the Wild

Open3D: A Modern Library for 3D Data Processing

PointFusion: Deep Sensor Fusion for 3D Bounding Box Estimation

Real-Time Seamless Single Shot 6D Object Pose Prediction

Frustum PointNets for 3D Object Detection from RGB-D Data

VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection

Adversarial Learning of Structure-Aware Fully Convolutional Networks for Landmark Localization

A Coarse-Fine Network for Keypoint Localization

Improving Landmark Localization with Semi-Supervised Learning

6-DoF object pose from semantic keypoints

Real-Time Perception Meets Reactive Motion Generation

Probabilistic Articulated Real-Time Tracking for Robot Manipulation

Learning to Track at 100 FPS with Deep Regression Networks

Depth-based object tracking using a Robust Gaussian Filter

Deep Tracking: Seeing Beyond Seeing Using Recurrent Neural Networks

ShapeNet: An Information-Rich 3D Model Repository

3D Object Proposals for Accurate Object Class Detection

You Only Look Once: Unified, Real-Time Object Detection

Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

Viewpoints and keypoints

DART: Dense Articulated Real-Time Tracking

Microsoft COCO: Common Objects in Context

Probabilistic object tracking using a range camera

Vision meets robotics: The KITTI dataset

Tracking-Learning-Detection

Representation Learning: A Review and New Perspectives

Are we ready for autonomous driving? The KITTI vision benchmark suite

ORB: An efficient alternative to SIFT or SURF

The MOPED framework: Object recognition and pose estimation for manipulation

6-DoF model-based tracking of arbitrarily shaped 3D objects

BRIEF: Binary Robust Independent Elementary Features

Real-time 3D model-based tracking using edge and keypoint features for robotic manipulation

Recognizing linked events: Searching the space of feasible explanations

Object tracking: A survey

Keypoint recognition using randomized trees

Machine Learning for High-Speed Corner Detection

SURF: Speeded Up Robust Features

Distinctive Image Features from Scale-Invariant Keypoints

Point matching as a classification problem for fast and robust object pose estimation

Least-Squares Fitting of Two 3-D Point Sets

Generalized-ICP

ORB: an efﬁcient alternative to SIFT or SURF

Ieee Transactions on Pattern Analysis and Machine Intelligence High-speed Tracking with Kernelized Correlation Filters

Field of Study

Journal Information

Name

Page

Venue Information

Name

Type

URL

Alternate Names