Convolutional-Recursive Deep Learning for 3D Object Classification

By Richard Socher, Brody Huval, Bharath Putta Bath, Christopher D. Manning, Andrew Y. Ng. 2012

648 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.

In 3D Vision & Pose

Citations in the corpus

Volumetric and Multi-view CNNs for Object Classification on 3D Data (2016, background): “A very first approach combines convolutional and recursive neural networks for learning features and classifying RGB-D images [27].”
Vehicle Detection from 3D Lidar Using Fully Convolutional Network (2016, methodology): “For example on RGBD data, one common aspect is to treat the depthmaps as image channels and use 2D CNN for classiﬁcation or detection [9, 24, 26].”
Deep Sliding Shapes for Amodal 3D Object Detection in RGB-D Images (2015, methodology): “2D Object Detector in RGB-D Images 2D object detection approaches for RGB-D images treat depth as extra channel(s) appended to the color images, using handcrafted features [9], sparse coding [2, 3], or recursive neural networks [23].”
Semantic Image Segmentation with Task-Specific Edge Detection Using CNNs and a Discriminatively Trained Domain Transform (2015, background): “Some recent work has attempted to model spatial long range dependency with recurrent networks for computer vision tasks [17, 41, 35, 4, 43].”
Understanding Deep Architectures using a Recursive Convolutional Network (2013, methodology): “More recently [25], they used a convolutional network in a separate stage to first learn features on RGB-Depth data, prior to hierarchical merging.”
Deeply-Recursive Convolutional Network for Image Super-Resolution (2015, methodology): “[25] used a convolutional network in a separate stage to first learn features on RGB-Depth data, prior to hierarchical merging.”
FPNN: Field Probing Neural Networks for 3D Data (2016, methodology): “5D data can be represented as multiple channel images, and processed by 2D CNNs [26, 10, 8].”
Orientation-boosted Voxel Nets for 3D Object Recognition (2016, neutral): “ShapeNet [1] 86 70 92 30 50 50 62 73 25 40 Recursive D [23] 0 73 81 10 47 22 34 48 41 20 Recursive D+C [23] 0 74 69 17 47 39 47 60 44 50 Our Baseline 76 82 90 23 73 37 71 74 64 89 ORION Basic 76 83 92 26 71 35 73 73 62 93 ORION Fusion 76 82 91 25 68 40 72 73 63 91 ORION F2C 76 82 91 26 67 35 74 72 6”
Indoor Semantic Segmentation using depth information (2013, background): “Now that depth maps are easy to acquire, deep learning approachs started to be considered for improving object recognition [20].”
Convolutional Neural Networks for Steady Flow Approximation (2016, methodology): “Convolutional neural networks have been proven successful in learning geometry representations [27, 13].”
Incremental On-Line Learning of Object Classes using a Combination of Self-Organizing Incremental Neural Networks and Deep Convolutional Neural Networks* (2016, positive): “Some of these approaches have been tailored for the object recognition task by using a combination of RGB and depth information ([4], [5]) and ﬁne-tuning or otherwise using pre-trained CNNs ([6], [7], [8]).”
The Role of Syntax in Vector Space Models of Compositional Semantics (2013, methodology): “Previously, RAE have successfully been applied to a number of tasks including sentiment analysis, paraphrase (Blacoe and Lapata, 2012; Socher et al., 2011b; Socher et al., 2012a).”
Dynamic Scene Deblurring With Parameter Selective Sharing and Nested Skip Connections (2019, methodology): “[28] used a CNN to first learn translational invariance features and applied the same network recursively to learn hierarchical feature representations in a tree structure.”
Classification With an Edge: Improving Semantic Image Segmentation with Boundary Detection (2016, methodology): “Socher et al. (2012) start their semantic segmentation pipeline with a single convolution and pooling layer.”
Classification and Segmentation of Satellite Orthoimagery Using Convolutional Neural Networks (2016, background): “Recently, CNNs have been shown to be successful in object recognition [17,18], object detection [19], scene parsing [20,21] and scene classification [22–24].”
3D ShapeNets: A deep representation for volumetric shapes (2014, methodology): “1,390 instances are used to train the algorithm of [29] and perform our discriminative fine-tuning, while the remaining 495 instances are used for testing all five methods.”