6D位姿估计 2021顶会论文汇总(持续更新~)

文章包括CVPR2021,ICCV2021,WACV2021
有其他文章欢迎提出意见

ICCV2021

  1. StereOBJ-1M: Large-scale Stereo Image Dataset for 6D Object Pose Estimation ——paper

  2. SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation ——paper

  3. CAPTRA: Category-level Pose Tracking for Rigid and Articulated Objects from Point Clouds ——paper

  4. RePOSE: Fast 6D Object Pose Refinement via Deep Texture Rendering —— paper

CVPR2021

  1. GDR-Net: Geometry-Guided Direct Regression Network for Monocular 6D Object Pose Estimation ——paper
  2. DSC-PoseNet: Learning 6DoF Object Pose Estimation via Dual-scale Consistency —— paper
  3. StablePose: Learning 6D Object Poses from Geometrically Stable Patches —— paper
  4. FS-Net: Fast Shapebased Network for Category-Level 6D Object Pose Estimation with Decoupled Rotation Mechanism —— paper

WACV2021

  1. A Pose Proposal and Refinement Network for Better 6D Object Pose Estimation ——paper

ICRA2021

  1. CloudAAE: Learning 6D Object Pose Regression with On-line Data Synthesis on Point Clouds —— paper

  2. RGB Matters: Learning 7-DoF Grasp Poses on Monocular RGBD Images —— paper

  3. Investigations on Output Parameterizations of Neural Networks for Single Shot 6D Object Pose Estimation —— paper

  4. Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenes —— paper

  5. ParametricNet: 6DoF Pose Estimation Network for Parametric Shapes in Stacked Scenarios —— paper

  6. REDE: End-to-End Object 6D Pose Robust Estimation Using Differentiable Outliers Elimination —— paper

IROS2021

  1. BundleTrack: 6D Pose Tracking for Novel Objects without Instance or Category-Level 3D Models ——paper
  2. KDFNet: Learning Keypoint Distance Field for 6D Object Pose Estimation ——paper

ACM MM 21

  1. GCCN: Geometric Constraint Coattention Network for 6D Object Pose Estimation —— paper

NeurIPS2021

  1. Leveraging SE(3) Equivariance for Self-supervised Category-Level Object Pose Estimation from Point Clouds ——paper

AAAI 2021

  1. SD-Pose: Semantic Decomposition for Cross-Domain 6D Object Pose Estimation ——paper

SCI

  1. 6D pose estimation with combined deep learning and 3D vision techniques for a fast and accurate object grasping —— paper

Arxiv

  • MPF6D: Masked Pyramid Fusion 6D Pose Estimation ——paper

  • 2021/12/09 更新
### 关于6D位姿估计的研究论文 #### 什么是6D位姿估计6D位姿估计是指确定对象在三维空间中的位置(3自由度)和方向(3自由度),总共涉及六个维度的信息。这项技术广泛应用于机器人视觉、增强现实(AR)、自动驾驶等领域[^1]。 #### 推荐的6D位姿估计相关论文 以下是几篇重要的6D位姿估计研究论文: 1. **《6D Object Pose from Semantic Keypoints》** 这篇文章探讨了一种基于语义关键点的方法来进行6D目标姿态估计。该方法利用深度学习模型提取目标的关键特征点,并通过几何约束计算其精确的姿态[^2]。 2. **《PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes》** 提出了PoseCNN框架,它能够实现在杂乱场景下的鲁棒6D物体姿态估计。此方法结合了卷积神经网络(CNNs)的强大表征能力和传统几何优化手段的优点[^3]。 3. **《DeepIM: Deep Iterative Matching for 6D Pose Estimation》** 文章介绍了一个名为DeepIM的新颖架构,旨在解决复杂光照条件下的高精度6D姿态估计问题。DeepIM采用迭代匹配策略逐步细化初始猜测的位置与角度参数[^4]。 4. **《SSD-6D: Making RGB-Based Dense 6D Object Pose Estimation and Correspondence Learning Practical》** SSD-6D是一种高效的端到端解决方案,能够在标准RGB图像基础上完成密集的目标定位任务。它的设计特别适合实时应用场合,在速度和准确性之间取得了良好平衡[^5]。 5. **《MVPosNet: Multi-view Consistent Point Cloud Completion for Accurate 6D Pose Estimation of Occluded Objects》** MVPosNet专注于处理遮挡情况下的精准6D姿态预测挑战。借助多视角一致性原则指导点云补全过程,从而提升最终结果的质量[^6]。 ```python import numpy as np def calculate_6d_pose(keypoints, model_points): """ 使用语义关键点估算6D姿势 参数: keypoints (list): 图像上的检测到的关键点坐标列表. model_points (list): 对应的3D模型点坐标列表. 返回: tuple: 包含旋转矩阵和平移向量的结果元组(R,t). """ R, t = estimate_transformation(keypoints, model_points) return R, t def estimate_transformation(src_pts, dst_pts): # 实现PnP算法或其他适配器... pass ```
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包

打赏作者

wakaka_Yu

你的鼓励将是我创作的最大动力

¥1 ¥2 ¥4 ¥6 ¥10 ¥20
扫码支付:¥1
获取中
扫码支付

您的余额不足,请更换扫码支付或充值

打赏作者

实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值