A mixed-reality dataset for category-level 6D pose and size estimation of hand-occluded containers (2022-11-18T00:00:00.000000Z)

TL;DR

This paper presents a mixed-reality dataset of hand-occluded containers for category-level 6D object pose and size estimation, and re-train and test an existing model for 6D objects pose estimation on this dataset.

Abstract

Estimating the 6D pose and size of household containers is challenging due to large intra-class variations in the object properties, such as shape, size, appearance, and transparency. The task is made more difficult when these objects are held and manipulated by a person due to varying degrees of hand occlusions caused by the type of grasps and by the viewpoint of the camera observing the person holding the object. In this paper, we present a mixed-reality dataset of hand-occluded containers for category-level 6D object pose and size estimation. The dataset consists of 138,240 images of rendered hands and forearms holding 48 synthetic objects, split into 3 grasp categories over 30 real backgrounds. We re-train and test an existing model for 6D object pose estimation on our mixed-reality dataset. We discuss the impact of the use of this dataset in improving the task of 6D pose and size estimation.

Authors

A. Cavallaro

2 papers

Xavier Weber

1 papers

A. Xompero

1 papers

TL;DR

Abstract

Authors

References33 items

Audio-Visual Object Classification for Human-Robot Collaboration

The CORSMAL benchmark for the prediction of the properties of containers

Objectron: A Large Scale Dataset of Object-Centric Videos in the Wild with Pose Annotations

CosyPose: Consistent multi-view multi-object 6D pose estimation

Object Handovers: A Review for Robotics

Rescaling Egocentric Vision: Collection, Pipeline and Challenges for EPIC-KITCHENS-100

Benchmark for Human-to-Robot Handovers of Unseen Containers With Unknown Filling

KeyPose: Multi-View 3D Labeling and Keypoint Estimation for Transparent Objects

HOnnotate: A Method for 3D Annotation of Hand and Object Poses

Learning Joint Reconstruction of Hands and Manipulated Objects

HomebrewedDB: RGB-D Dataset for 6D Pose Estimation of 3D Objects

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

Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects

BOP: Benchmark for 6D Object Pose Estimation

Embodied hands

PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes

Cut, Paste and Learn: Surprisingly Easy Synthesis for Instance Detection

Mask R-CNN

Deep Residual Learning for Image Recognition

Semantically-enriched 3D models for common-sense knowledge

Fast R-CNN

Microsoft COCO: Common Objects in Context

Model Based Training, Detection and Pose Estimation of Texture-Less 3D Objects in Heavily Cluttered Scenes

Indoor Segmentation and Support Inference from RGBD Images

Graspit! A versatile simulator for robotic grasping

Least-Squares Estimation of Transformation Parameters Between Two Point Patterns

Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography

ShAPO: Implicit Representations for Multi-Object Shape, Appearance, and Pose Optimization

“The Open Images Dataset V4: Uni-ﬁed image classiﬁcation, object detection, and visual relationship detection at scale,”

Blender -a 3D modelling and rendering package

Blender Online Community

labelme: Image Polygonal Annotation with Python

Noname manuscript No. (will be inserted by the editor) EPnP: An Accurate O(n) Solution to the PnP Problem

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names