Geodesic Object Proposals

By Philipp Krahenbuhl, Vladlen Koltun. 2014

396 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

Shape-aware Instance Segmentation (2016, methodology): “Beyond instance-level semantic segmentation, many methods have been proposed to generate class-agnostic region proposals [1, 31, 18].”
Training Region-Based Object Detectors with Online Hard Example Mining (2016, methodology): “R-CNN, in contrast, uses region proposals [1, 2, 3, 4, 6, 10, 18, 32, 34], a method that was made popular by the selective search algorithm [32].”
Learning to Segment Object Candidates (2015, methodology): “These approaches usually fall into three categories: (1) objectness scoring [1, 34], in which proposals are extracted by measuring the objectness score of bounding boxes, (2) seed segmentation [14, 16, 17], where models start with multiple seed regions and generate separate foreground-background seg”
What Makes for Effective Detection Proposals? (2015, neutral): “Geodesic [22] Grouping @ @ 1 $ $ $ $ $ $”
Joint Discovery of Object States and Manipulation Actions (2017, methodology): “In detail, we use geodesic object proposals [27] and set a relatively low object detection threshold (0.”
Object-Proposal Evaluation Protocol is ‘Gameable’ (2015, background): “More recently, object proposals are increasingly viewed as detection proposals [1, 8, 11, 42] where the goal is to improve the object detection pipeline, focusing on a chosen set of object classes (e.”
Scalable, High-Quality Object Detection (2014, methodology): “l imaging could be replaced by a more general methodology, improving their quality and simplifying their pipelines as well. Given the recent success of novel sophisticated proposal generation methods [1,22,24,14,3,10,2,11], the possibility of directly learned methods overtaking engineered methods mi”
BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentation (2015, methodology): “In Table 3, we compare the effects of various region proposals on our method: Selective Search (SS) [31], Geodesic Object Proposals (GOP) [17], and MCG [2].”
Category-specific object reconstruction from a single image (2014, background): “At the core of our approach are novel deformable 3D models that can be learned from 2D annotations available in existing object detection datasets, that can be driven by noisy automatic object segmentations and which we complement with a bottom-up module for recovering high-frequency shape details.”
LSDA: Large Scale Detection through Adaptation (2014, methodology): “We have therefore also implemented and released a version of our algorithm running with fast region proposals [27] on a spatial pyramid pooling network [28], reducing our detection time down to half a second per image (from 4s per image) with nearly the same performance.”
Modeling Relationships in Referential Expressions with Compositional Modular Networks (2016, neutral): “via object proposal methods like [28, 4, 12, 35]).”
Attend Refine Repeat: Active Box Proposal Generation via In-Out Localization (2016, background): “Several approaches have been proposed in the literature for this task [1, 2, 5, 6, 8, 18, 25, 26, 29, 40, 44].”
Multiscale Combinatorial Grouping for Image Segmentation and Object Proposal Generation (2015, background): “The most recent wave of object proposal algorithms is represented by [13], [14], and [15], which all keep the quality of the seminal proposal works while improving the speed considerably.”
Self-taught object localization with deep networks (2014, methodology): “Objectness methods [2, 6, 30, 4, 3, 16, 36, 25, 21, 26, 32, 14] aim at generating bounding boxes that yield high recall for a high number of candidates, i.”
Recurrent Pixel Embedding for Instance Grouping (2017, methodology): “We compare against the most recent techniques including POISE [38], LPO [51], CPMC [8], GOP [50], SeSe [82], GLS [71], RIGOR [37].”
DeepBox: Learning Objectness with Convolutional Networks (2015, methodology): “A similar idea is used in GOP [20] which uses a much faster geodesic based segmentation instead of graph cuts.”