Author manuscript, published in "International Conference on Computer Vision (ICCV) (2013)" Segmentation Driven Object Detection with Fisher Vectors

2013

Citations in the corpus

• Unsupervised Visual Representation Learning by Context Prediction (2015, neutral): “le bus car cat chair cow table dog horse mbike person plant sheep sofa train tv mAP DPM-v5[17] 33.2 60.3 10.2 16.1 27.3 54.3 58.2 23.0 20.0 24.1 26.7 12.7 58.1 48.2 43.2 12.0 21.1 36.1 46.0 43.5 33.7 [8] w/o context 52.6 52.6 19.2 25.4 18.7 47.3 56.9 42.1 16.6 41.4 41.9 27.7 47.9 51.5 29.9 20.0 41.1”
• What Makes for Effective Detection Proposals? (2015, background): “In addition to allowing for use of more sophisticated classifiers, the use of detection proposals alters the data distribution that the classifiers handle.”
• 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).”
• Weakly Supervised Object Localization with Multi-Fold Multiple Instance Learning (2015, positive): “The FV-based representation is based on our previous work [13] for fully supervised detection.”
• AttentionNet: Aggregating Weak Directions for Accurate Object Detection (2015, methodology): “However, previous methods depends on these object proposals [6, 5, 12, 17, 21] grade these windows by SVM scores only, without an intrinsic model to carefully handle the problem.”
• Large-Scale Weakly Supervised Object Localization via Latent Category Learning (2015, methodology): “In this paper, we use a recent segmentation based region proposal named Selective Search [38], which uses multiple low-level cues, hierarchical segmentation and various grouping strategies to generate regions in which objects are likely to be found [39], [40].”
• Regionlets for Generic Object Detection (2013, methodology): “Performing object detection with object proposals has been attracting increasing attentions [ 31], [32], [33], [34], [35], [8].”
• Spatio-temporal Object Detection Proposals (2014, unknown): “Fisher vectors [8], max-pooling regionlets [42], and convolutional networks [17].”
• Non-maximum Suppression for Object Detection by Passing Messages Between Windows (2014, background): “Most current object category detection pipelines [12–14], but also generic object detection ones [7], use such a greedy procedure.”
• Towards Unified Object Detection and Semantic Segmentation (2014, unknown): “In addition, some ambiguous object hypotheses may be difficult to classify from the information within the window/segment alone, but contextual information, such as local context around each object hypothesis and global image-level context, can help [28,34,11].”
• Second-order constrained parametric proposals and sequential search-based structured prediction for semantic segmentation in RGB-D images (2015, methodology): “For intensity images, significant progress has been achieved recently through work performed in association with the VOC Semantic Segmentation challenges[11], where high performing methods for boundary detection[36, 31], feature description and nonlinear feature maps[4, 39, 37, 46, 33, 6], image seg”
• Multi-fold MIL Training for Weakly Supervised Object Localization (2014, unknown): “To represent (tentative) detection windows, we use the high-dimensional Fisher vector ( FV ) image representation [24], which has shown to yield state-of-the-art performance for image classification and object detection [6, 7, 9].”
• Local Alignments for Fine-Grained Categorization (2014, background): “Furthermore, performing segmentation for all the bounding box candidates returned by state-of-the-art object detectors, like (Uijlings et al. 2013; Cinbis et al. 2013), would be computationally challenging.”
• Contextual Priming and Feedback for Faster R-CNN (2016, methodology): “In fact, there has been a lot of interest in using segmentation in tandem with detection [30, 31, 37, 38]; e.”
• Semantic Image Interpretation - Integration of Numerical Data and Logical Knowledge for Cognitive Vision (2018, unknown): “These take the input proposals and classify (or discard) them with some labels [46, 114, 104, 45, 101, 21, 58, 36, 37].”
• DeepProposal: Hunting Objects by Cascading Deep Convolutional Layers (2015, background): “Most of the recent state-ofthe-art detection methods [6, 12, 14, 25] are based on such proposals.”
• Model recommendation: Generating object detectors from few samples (2015, methodology): “For example, flagship techniques such as Deformable Part Models (DPMs) [14] via sliding-window fashion are best suitable for semi-rigid objects, while Exemplar-SVMs [33] with simple linear SVMs are highly category and instance specific; the recent top performing detection systems instead favor botto”
• A discriminative CNN video representation for event detection (2014, methodology): “With further compression with a lossless technique such as Blosc(3)[8], we can store the features of the whole collection in less than 1 GB, which can be read by a normal SSD disk in a few seconds.”
• Bags of Spacetime Energies for Dynamic Scene Recognition (2014, background): “Recently, the Fisher vector representation [19] has shown state-of-the-art results for a variety of visual tasks (e.g., [5, 14, 18, 23]); in contrast, here it is found that locality constrained linear coding [27] performs particularly well for dynamic scenes.”