Video Segmentation with Superpixels

By Fabio Galasso, Roberto Cipolla, Bernt Schiele. 2012

117 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

COCO-Stuff: Thing and Stuff Classes in Context (2016, methodology): “Several works use interactive segmentation methods [42, 57, 10] to speedup annotation; others annotate superpixels [61, 22, 40].”
Video Pop-up: Monocular 3D Reconstruction of Dynamic Scenes (2014, background): “On the other hand, superpixel [13] and supervoxel [41] methods for video segmentation can produce high quality video over-segmentations that respect object boundaries, are temporally consistent and are aligned with objects.”
Learning Must-Link Constraints for Video Segmentation Based on Spectral Clustering (2014, methodology): “Figure 1 illustrates the advantages of learning must-links: superpixel-based techniques [18] build spectral graphs on fine superpixels, Figure 1(b); by contrast, we propose to build graphs merging superpixels based on learned must-link constraints, Figure 1(c).”
A Unified Video Segmentation Benchmark: Annotation, Metrics and Analysis (2013, methodology): “[10] provide coarse-to-fine video segmentation and could be additionally employed for the task.”
SegFlow: Joint Learning for Video Object Segmentation and Optical Flow (2017, methodology): “Several methods have been proposed to generate object segmentation via saliency [31, 11, 42], optical flow [4, 28] or superpixel [17, 46, 13].”
Spectral Graph Reduction for Efficient Image and Streaming Video Segmentation (2014, methodology): “The graph is either constructed on pre-computed superpixels/supervoxels [27, 12, 16, 7] which are then clustered, or the clustering is (a) Original graph (b) Equivalent graph Figure 1.”
FusionSeg: Learning to Combine Motion and Appearance for Fully Automatic Segmentation of Generic Objects in Videos (2017, methodology): “In recent years, video object segmentation has received significant attention, with great progress on fully automatic algorithms [15, 56, 13, 23, 30, 58, 24, 35, 10], propagation methods [40, 47, 11, 50, 19, 53], and interactive methods [51, 26, 2, 42].”
Semi-automatic video object segmentation by advanced manipulation of segmentation hierarchies (2015, methodology): “STATE OF THE ART Video segmentation techniques [9], [10], [11], [12] take a video as input and output a partition of the frames into a set of regions aiming at being as temporally consistent as possible, and at regions between partitions being as connected as possible.”
Spatio-Temporal Moving Object Proposals (2014, background): “However, combining ﬂow and image boundaries has so far only shown moderate success [13, 38].”
Click Carving: Segmenting Objects in Video with Point Clicks (2016, methodology): “ds use no human input, and typically produce an over-segmentation of the video that is useful for mid-level grouping. Supervoxel methods ﬁnd space-time blobs cohesive in color and/or motion [1], [2], [3], while point trajectory approaches ﬁnd consistent motion threads beyond optical ﬂow [19], [20]. ”
Learning to segment moving objects in videos (2014, background): “In contrast, many researchers have tried combining optical ﬂow with static boundaries in order to improve boundary detection [34,37], with only moderate success so far [11, 37].”