Object detection and segmentation from joint embedding of parts and pixels

By Michael Maire, Stella X. Yu, Pietro Perona. 2011

65 citations recorded by the source corpus.

Subfields: 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 Computer Vision and Pattern Recognition

Citations in the corpus

Unified Perceptual Parsing for Scene Understanding (2018, background): “For example, a number of previous research works focus on the combination of recognition and segmentation [19,20,21].”
The Cityscapes Dataset for Semantic Urban Scene Understanding (2016, methodology): “Following this line, [9, 37, 40, 63, 81] propose different approaches that combine the strengths of CNNs and Conditional Random Fields (CRFs).”
UberNet: Training a Universal Convolutional Neural Network for Low-, Mid-, and High-Level Vision Using Diverse Datasets and Limited Memory (2016, background): “Finally, the particular motivation for this research has been the interest in studying the synergy between different visual tasks (e.g. the long-standing problem of combining segmentation and recognition [9, 47, 50, 53, 68, 72, 99]), so this work can be understood as a ﬁrst step in this direction.”
Iterative Instance Segmentation (2015, methodology): “Another alternative is to use eigenvectors obtained from normalized cuts as an embedding for pixels [23, 22].”
segDeepM: Exploiting segmentation and context in deep neural networks for object detection (2015, methodology): “In [1], object segmentations are found by aligning the masks obtained from Poselets [1, 17].”
Associative Embedding: End-to-End Learning for Joint Detection and Grouping (2016, methodology): “Prior works typically use a two stage pipeline [29], detecting basic visual units (patches, superpixels, parts, etc.”
Joint Object and Part Segmentation Using Deep Learned Potentials (2015, background): “Poselets [5] are proposed to model the local human parts through 2D projections from 3D data, which can be used as robust representation for both detection [20] and segmentation [29, 1].”
Symbiotic Segmentation and Part Localization for Fine-Grained Categorization (2013, background): “In the poselet line of research [6] the detectors are for parts, rather than for entire categories, but again the poseletdetectors can predict foreground masks for object category detection and segmentation [9, 20].”
Interestingness Prediction by Robust Learning to Rank (2014, methodology): “By aggregating pairwise local rankings into a global ranking, methods such as Huber-LASSO [46,18] have the potential to be robust against local ranking noise [5,31].”
Affinity CNN: Learning Pixel-Centric Pairwise Relations for Figure/Ground Embedding (2015, background): “For example, consistent definitions for figure/ground include true depth ordering as in [9], object class-specific foreground/background separation as in [24], and boundary ownership or occlusion as in [27, 13, 22].”
Semantic segmentation using regions and parts (2012, methodology): “[24] solve the combined problem of detection and segmentation of people by incorporating outputs of top down part-detectors in a generalized eigen-problem and obtaining parts and pixel groupings.”