Shape and Appearance Context Modeling

By Xiaogang Wang, Gianfranco Doretto, Thomas B. Sebastian, Jens Rittscher, Peter Henry Tu. IEEE International Conference on Computer Vision, 2007

556 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/59c8447ef5fe6b4057148553ea3c208447876145

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

1. Xiaogang Wang
2. Yi Yang
3. Qi Tian
4. Luc Van Gool
5. Wanli Ouyang
6. Liang Zheng
7. Dacheng Tao
8. Shaogang Gong

Citations in the corpus

• Large scale metric learning from equivalence constraints (2012, methodology): “To indicate the performance of the various algorithms we report Cumulative Matching Characteristic (CMC) curves [18].”
• Jointly Attentive Spatial-Temporal Pooling Networks for Video-Based Person Re-identification (2017, background): “A majority of recent works mainly develop their solutions from two aspects: extracting reliable feature representations [28, 6, 11, 19, 15, 38] or learning a robust distance metric [14, 41, 2, 30, 36, 37, 21, 13].”
• Person re-identification by symmetry-driven accumulation of local features (2010, methodology): “As to single-shot approaches, in [20] the method consists in segmenting a pedestrian image into regions, and registering their color spatial relationship into a co-occurrence matrix.”
• Person Re-identification by Descriptive and Discriminative Classification (2011, methodology): “These approaches can mainly be subdivided in two groups: (a) methods which employ a representation of descriptive statistics of the human appearance (using hand crafted features) [4,6,17,22] and (b) approaches that are based on discriminative learning [3,8,13,16,18].”
• Similarity Learning with Spatial Constraints for Person Re-identification (2016, methodology): “As pedestrians have relative stable geometric structure, spatial constraints have been widely adopted for person image representation [29, 10, 36, 17, 20, 30, 26, 34].”
• Salient Color Names for Person Re-identification (2014, methodology): “All the results are shown in form of Cumulated Matching Characteristic (CMC) curve [25].”
• Learning to rank in person re-identification with metric ensembles (2015, background): “Existing studies have tried to address the above problem by seeking a combination of robust and distinctive feature representation of person’s appearance, ranging from color histogram [12], spatial co-occurrence representation [40], LBP [44], color SIFT [46], etc.”
• Viewpoint Invariant Pedestrian Recognition with an Ensemble of Localized Features (2008, methodology): “If the problem is limited to a frontal viewpoint then one could fit a triangular graph model [6] or part based model [7] to account for pose change.”
• PersonNet: Person Re-identification with Deep Convolutional Neural Networks (2016, background): “Many recent studies on person re-identification attempt to generate robust feature representation which is discriminative and robust for describing a pedestrian’s appearance under various changes and conditions [2], [4], [7], [11], [28]–[30].”
• Deep feature learning with relative distance comparison for person re-identification (2015, methodology): “The features used in person reidentification range from the color histogram [2], spatial cooccurrence representation model [3], attributes model [4] to combination of multiple features [2,5].”
• Intelligent multi-camera video surveillance: A review (2013, methodology): “Histogram of Oriented Gradients (HOG) (Dalal and Triggs, 2005; Wang et al., 2007; Schwartz and Davis, 2009) characterizes local shapes by capturing edges and gradient structures.”
• Cross Dataset Person Re-identification (2014, unknown): “Another type of person re-identification methods try to tackle the problem by seeking feature representations which are both distinctive and stable for describing the appearance[14–17].”
• Person Re-identification by Multi-Channel Parts-Based CNN with Improved Triplet Loss Function (2016, methodology): “Once the feature extraction is completed, these methods usually choose a standard distance measure such as l1-norm based distance (L1norm) [40], Bhattacharyya distance (Bhat)[14], or Mahalanobis distance (Maha)[33] to determine the similarity between pairs.”
• Person re-identification by probabilistic relative distance comparison (2011, background): “sure such as l1-norm [14], l2-norm based distance [8], or Bahattacharyya distance [7].”
• Partial Person Re-Identification (2015, background): “Person re-identification (re-id) has been studied extensively in the past five years [5, 30, 4, 37, 10, 20, 36].”
• Deep Learning for Person Re-Identification: A Survey and Outlook (2020, methodology): “Open-set : In the pedestrian retrieval stage (Step 5), most existing closed-world person Re-ID works assume that the query must occur in the gallery set by calculating the CMC [58] and mAP [5].”
• DeepReID: Deep Filter Pairing Neural Network for Person Re-identification (2014, background): “Global features characterize the distributions of color and texture with the histograms of visual words [4, 43].”
• Person Re-identification by Salience Matching (2013, methodology): “All the quantitative results are reported in standard Cumulated Matching Characteristics (CMC) curves [20].”
• Human Reidentification with Transferred Metric Learning (2012, background): “Many visual features of characterizing color [5, 6], shape [7, 8] and texture [9–11] of objects have been proposed.”
• Person Re-Identification by Support Vector Ranking (2010, methodology): “To evaluate comparatively all six re-identification methods, we display the cumulative matching characteristic (CMC) curve [13], which is based on the ranking of each of the gallery image with respect to the probe, thus resulting in the expectation of the”