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

By Stefan Hinterstoißer, Vincent Lepetit, Slobodan Ilic, Stefan Holzer, Kurt Konolige, Gary R. Bradski, Nassir Navab. 2012

1,493 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

RGB-D datasets using microsoft kinect or similar sensors: a survey (2016, methodology): “More details about this dataset can be found in [36, 37].”
OctNetFusion: Learning Depth Fusion from Data (2017, methodology): “Shape Completion: If exact 3D models are available, missing surfaces can be completed by detecting the objects and fitting 3D models to the observations [3, 21].”
Unsupervised Pixel-Level Domain Adaptation with Generative Adversarial Networks (2016, methodology): “Table 2 reports the mean angle the object would need to be rotated (on a fixed 3D axis) to move from predicted to ground truth pose [22].”
Domain Separation Networks (2016, methodology): “2 we report the mean angle the object would need to be rotated (on a fixed 3D axis) to move from the predicted to the ground truth pose [12].”
SE3-nets: Learning rigid body motion using deep neural networks (2016, methodology): “We use 11 household objects from the LineMOD dataset [27].”
PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes (2017, methodology): “With the emergence of depth cameras, several methods have been proposed for recognizing texture-less objects using RGB-D data [6], [7], [8], [9].”
BOP Challenge 2020 on 6D Object Localization (2020, methodology): “tasets can be downloaded from: bop.felk.cvut.cz/datasets. BOP Challenge 2020 on 6D Object Localization 7 Train. im. Val im. Test im. Test inst. Dataset Core Objects Real PBR Real All Used All Used LM [18] 15 { 50000 { 18273 3000 18273 3000 LM-O [3] 8 { 50000 { 1214 200 9038 1445 T-LESS [23] 30 37584”
DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion (2019, methodology): “Table 3 compares our method with previous RGB methods with depth refinement(ICP) (results from [31, 34]) on the ADD metric [13].”
DeepIM: Deep Iterative Matching for 6D Pose Estimation (2018, methodology): “For symmetric objects, we use the closest point distance in computing the average distance for 6D pose evaluation as in Hinterstoisser et al. (2012b): ADD-S = 1 m ∑ x1∈M min x2∈M ‖(Rx1 + t) − (R̂x2 + t̂)‖2.”
3D-FUTURE: 3D Furniture Shape with TextURE (2020, methodology): “Recentworks thus employ RGB-D sensors and deep learning to improve keypoints detection or directly predict 6DoF pose from images (Kehl et al. 2016; Brachmann et al. 2014; Bo et al. 2014; Hinterstoisser et al. 2012; Xiang et al. 2017; Peng et al. 2019; Song et al. 2020; Tekin et al. 2018; Rad and Lep”
DRIT++: Diverse Image-to-Image Translation via Disentangled Representations (2019, methodology): “rthermore, we demonstrate the potential application of unsupervised domain adaptation. On the tasks of adapting domains from MNIST [24] to MNISTM [12] and Synthetic Cropped LineMod to Cropped LineMod [15,43], we show competitive performance against state-of-the-art domain adaptation methods. We make”
GraspNet-1Billion: A Large-Scale Benchmark for General Object Grasping (2020, background): “Firstly, the grasp poses have different representations including rectangle [36] and 6D pose [41] representation and are evaluated with different metrics [16, 14, 41] correspond-”
PVN3D: A Deep Point-Wise 3D Keypoints Voting Network for 6DoF Pose Estimation (2019, methodology): “The ADD(S)[19] AUC is computed in a similar way but calculate ADD distance for non-symmetric objects and ADD-S distance for symmetric objects.”
SUN RGB-D: A RGB-D scene understanding benchmark suite (2015, neutral): “There are many existing RGB-D datasets available [54] [47] [1] [25] [44] [60] [49] [45] [29] [66] [57] [52] [46] [16] [21] [73] [35] [67] [3] [41] [10] [63] [42] [64] [65] [48] [12] [33] [8] [50] [26] [6].”
Latent-Class Hough Forests for 3D Object Detection and Pose Estimation (2014, methodology): “State-of-the-art approaches in 3D detection and pose estimation [21, 36] avoid this issue by training just from 3D models of the target object.”
Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects (2018, methodology): “Since PoseCNN has already been shown to outperform the single-shot method of Tekin et al. [6], which itself beats the more complex networks of BB8 [4] and SSD-6D [22] on the standard LINEMOD [23] and Occluded-LINEMOD [24] datasets, the comparison of our method against PoseCNN is sufﬁcient to answer ”
Real-Time Seamless Single Shot 6D Object Pose Prediction (2017, methodology): “whose pose can only be computed up to one degree of rotational freedom, we modify slightly the metric as in [2, 8] and compute”
Few-Shot Object Detection and Viewpoint Estimation for Objects in the Wild (2020, methodology): “Deep-learning methods for viewpoint estimation follow roughly three different paths: direct estimation of Euler angles [55, 50, 33, 24, 62, 63], template-based matching [20, 32, 51], and keypoint detection relying on 3D bounding box corners [39, 52, 13, 34, 36] or semantic keypoints [35, 65].”
Image Matching from Handcrafted to Deep Features: A Survey (2020, background): “Modern attempts for object recognition also include some specifically handcrafted features (Dalal and Triggs 2005; Hinterstoisser et al. 2012) and, more recently, deep learning approaches (Wohlhart and Lepetit 2015).”