Semantic Segmentation with Second-Order Pooling

By João Carreira, Rui Caseiro, Jorge P. Batista, Cristian Sminchisescu. 2012

533 citations recorded by the source corpus.

Subfields: Artificial Intelligence, 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 Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Fast R-CNN (2015, methodology): “SegDeepM is trained on VOC12 train-val plus segmentation annotations; it is designed to boost R-CNN accuracy by using an MRF to reason over R-CNN detections and segmentations from the O 2 P [1] semantic-segmentation method.”
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation (2013, methodology): “The train images cannot be used for hard negative mining, because annotations are not exhaustive.”
Instance-Aware Semantic Segmentation via Multi-task Network Cascades (2015, neutral): “7 (%) time/img (s) OP [2] 25.”
Simultaneous Detection and Segmentation (2014, methodology): “SegDPM combined DPMs [13] with O2P [5] and achieved a 9 point boost over DPMs in classical object detection.”
DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs (2016, background): “Substantial improvements have been achieved by incorporating richer information from context [45] and structured prediction techniques [22], [26], [27], [46], but the performance of these systems has always been compromised by the limited expressive power of the features.”
Learning Open-World Object Proposals without Learning to Classify (2021, background): “They have become the integral pre-processing steps for many computer vision systems, including object detection [33,41,42,45,59,74], segmentation [3,8,13], object discovery [12, 16, 47], weakly supervised object detection [5, 21, 53], visual tracking [35, 63], content-aware retargeting [51], etc.”
Per-Pixel Classification is Not All You Need for Semantic Segmentation (2021, methodology): “In fact, before FCN [30], the best performing semantic segmentation methods like O2P [5] and SDS [20] used a mask classification formulation.”
Object-Proposal Evaluation Protocol is ‘Gameable’ (2015, background): “Types of Object Proposals: Object proposals can be broadly categorized into two categories: • Window scoring : In these methods, the space of all possible windows in an image is sampled to get a sub-set of the windows ( e.g ., via sliding window).”
Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs (2014, methodology): “A celebrated non-DCNN precursor to these works is the second order pooling method of (Carreira et al., 2012) which also assigns labels to the regions proposals delivered by (Carreira & Sminchisescu, 2012).”
Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation (2015, neutral): “O2P [2] 18.”
Feedforward semantic segmentation with zoom-out features (2014, methodology): “A similar approach is taken in [39], where multiple segmentations obtained from [4] are re-ranked using a discriminatively trained model.”
Lifting Object Detection Datasets into 3D (2015, methodology): “We also bypass segmentation, another yet incompletely solved vision problem despite much recent progress [9], [10], and rely on ground truth silhouettes as input to our dense reconstruction engine which is based on a novel visual hull based algorithm.”
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.”