Sparse distance learning for object recognition combining RGB and depth information

By Kevin Lai, Liefeng Bo, Xiaofeng Ren, Dieter Fox. 2011

230 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Learning human activities and object affordances from RGB-D videos (2012, methodology): “, 2012), semantic labeling of environments by modeling objects and their relations to other objects in the scene (Koppula et al., 2011; Lai et al., 2011b; Rosman and Ramamoorthy, 2011; Anand et al., 2012), developing frameworks for object recognition and pose estimation for manipulation (Collet et a”
Voting for Voting in Online Point Cloud Object Detection (2015, methodology): “[11] deploy the Kinect sensor to learn a sparse distance metric for recognising pre-segmented indoor objects with Group-Lasso regularisation.”
Data-Driven Grasp Synthesis—A Survey (2013, background): “Lai et al. [107] perform object category and instance recognition.”
3DNet: Large-scale object class recognition from CAD models (2012, methodology): “In the accompanied approach [12] multiple features are combined, trained, and evaluated on this dataset and the authors showed that shape together with color leads to improved object recognition.”
Pedestrian detection combining RGB and dense LIDAR data (2014, background): “Kinect provides a dense depth map registered with an RGB image (RGB-D) on which standard detection machinery can be applied and this has stimulated much recent research in object detection [7],[8] articulated pose prediction [9] and scene understanding [10].”
Learning Discriminative Representations from RGB-D Video Data (2013, background): “Such an example can be found in RGB-D object detection and recognition [Spinello and Arras, 2012; Lai et al., 2011].”
Histogram of Oriented Normal Vectors for Object Recognition with a Depth Sensor (2012, methodology): “We present the performance of 6 methods: HOG, HOGD, HONV, Depth Kernel Descriptors (DKD) [7], Hierarchical Kernel Descriptors (HKDES) [8] and Instance Distance Learning (IDL) [9].”
Human Activity Learning using Object Affordances from RGB-D Videos (2012, background): “Recently, with the availability of inexpensive RGB-D sensors, some works consider labeling and object recognition in 3D point-clouds[21, 22].”
Body Surface Context: A New Robust Feature for Action Recognition From Depth Videos (2014, methodology): “Then, we adopt the empirical converting equations used in [39]”
Sliding Shapes for 3D Object Detection in Depth Images (2014, background): “3D classification: Classification-based approaches [65–68, 51, 69–79] typically consider the whole object at the same time by extracting a holistic feature for the whole object and classifying the feature vector via a classifier.”
50 Years of object recognition: Directions forward (2013, background): “Within the context of more recent work, the success of Kinect-based systems [296, 297, 298, 299] demonstrates how combined active and passive sensing systems improve recognition.”
Contextually Guided Semantic Labeling and Search for 3D Point Clouds (2011, methodology): “y of synchronized videos of both color and depth obtained from RGB-D (Kinect-style) depth cameras, shifted the focus to making use of both visual as well as shape features for object detection (e.g., Lai et al., 2011a,b), 3D segmentation (e.g., Collet Romea et al., 2011), human pose estimation (Ly e”
Category Modeling from Just a Single Labeling: Use Depth Information to Guide the Learning of 2D Models (2013, background): “Recently, RGBD images made object detection much easier [14, 15, 16], and even the structure discovery [17] or segmentation of indoor environments [18, 19] produced object-level results.”
A Survey on Human Motion Analysis from Depth Data (2013, methodology): “Evaluations are conducted based on the Cornell 60 dataset [32] and a new dataset acquired by the authors, named Cornell 120.”
OmniTouch: wearable multitouch interaction everywhere (2011, background): “These simple features worked well in our prototype implementation given the small set of surfaces to distinguish, but a more general solution would require more sophisticated features (e.g., see [13] for depth-driven object recognition).”
Structure-Preserving Binary Representations for RGB-D Action Recognition (2016, background): “Methods in [36] and [37] simply optimize all available information in their algorithms for object detection and recognition respectively.”
Learning to place new objects in a scene (2012, methodology): “Instead of images or 3D models, Lai et al. (2011) proposed a learning algorithm for categorization using both RGB and multi-view point-cloud.”
Contextually guided semantic labeling and search for three-dimensional point clouds (2013, unknown): “…from RGB-D (Kinect-style) depth cameras shifted the focus to making use of both visual as well as shape features for object detection (e.g., Lai et al., 2011a,b), 3D segmentation (e.g., Collet Romea et al., 2011), human pose estimation (e.g., Ly et al., 2012) and human activity detection…”