Analyzing 3D Objects in Cluttered Images

By Mohsen Hejrati, Deva Ramanan. 2012

125 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Category-specific object reconstruction from a single image (2014, methodology): “Closer to our work, Hejrati and Ramanan [24] used NRSfM on a larger class (cars) but need a predictive detector to fill-in missing data (occluded keypoints) which we do not assume to have here.”
Single Image 3D Interpreter Network (2016, methodology): “First, traditional methods [43,9] that minimize the repro-jection error consider only one keypoint hypothesis, and is therefore not robust to noises in keypoint detection.”
Context as Supervisory Signal: Discovering Objects with Predictable Context (2014, methodology): “Finally, we demonstrate the quality of our discovered intra-cluster correspondences by applying our method to the problem of keypoint annotation using the car annotations on PASCAL 2011 from [35].”
Viewpoints and keypoints (2014, methodology): “Previous works [32, 15, 16] have also jointly tackled the problem of keypoint detection and pose estimation.”
Sparse Representation for 3D Shape Estimation: A Convex Relaxation Approach (2015, methodology): “[8], [5]), most of the failed cases were attributed to bad initialization.”
Aligning 3D models to RGB-D images of cluttered scenes (2015, background): “There is a lot of research on 3D scene understanding from a single RGB image [15, 29], and 3D object analysis [2, 16, 23, 30, 41].”
Modeling Occlusion by Discriminative AND-OR Structures (2013, background): “…estimate the visibilities of parts from image appearance, using assumptions that the visibility of a part is independent from other parts [30] [29] [13], is consistent with neighbouring parts [7] [9], or is consistent with its parent or child parts describing object appearance at different scales…”
Self-Supervised Learning of Geometrically Stable Features Through Probabilistic Introspection (2018, background): “Only a few works have considered keypoint detection for generic categories [11, 21, 38, 36].”
Discovery of Latent 3D Keypoints via End-to-end Geometric Reasoning (2018, background): “A body of work also exists for more generalized, albeit supervised, keypoint detection, e.g., [15, 61].”
Scene Parsing by Integrating Function, Geometry and Appearance Models (2013, background): “[24], Hejrati and Ramanan [14], Xiang and Savarese [22], Pepik et al.”
Data-driven 3D Voxel Patterns for object category recognition (2015, background): “5D object representation, object parts or features are connected across views [33, 31, 32, 20].”
Seeing 3D Chairs: Exemplar Part-Based 2D-3D Alignment Using a Large Dataset of CAD Models (2014, background): “In category-level recognition, recent work has explored recognition, alignment and fine pose estimation for cars and bicycles in outdoor scenes using low-dimensional parametric deformable 3D models combined with a relatively small number of learnt part detectors [20, 30, 41].”
FPM: Fine Pose Parts-Based Model with 3D CAD Models (2014, background): “Recently, there has been an increasing interest in utilizing 3D models for object detection and pose estimation [39,27,10,19,34].”
GS3D: An Efficient 3D Object Detection Framework for Autonomous Driving (2019, background): “Early works [26, 6, 29, 9, 5] use low-level feature or statistics analysis to tackle 3D recognition or recover tasks.”
Learning Category-Specific Deformable 3D Models for Object Reconstruction (2017, methodology): “Closer to our work, Hejrati and Ramanan [46] used NRSfM on a larger class (cars) but need a predictive detector to fill-in missing data (occluded keypoints) which we do not assume to have here.”
Inferring 3D Object Pose in RGB-D Images (2015, background): “There has been a lot of research on 3D scene understanding from a single RGB image [15, 28], and 3D object analysis [22, 2, 16, 39, 29].”
Is object localization for free? - Weakly-supervised learning with convolutional neural networks (2015, background): “…in visual recognition ranges from image-level labels (object/image classiﬁcation) [23], location and extent of objects in the form of bounding boxes (object detection) [18], to detailed object segmentation [5, 58, 24] or even predicting an approximate 3D pose and geometry of objects [ 26, 45].”