Crowd Counting Using Multiple Local Features

By David Ryan, Simon Denman, Clinton Fookes, Sridha Sridharan. 2009 Digital Image Computing: Techniques and Applications, 2009

321 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/d52e2861cf7e2db520a6f9bf75d4569eb55d398e

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

A Survey of Recent Advances in CNN-based Single Image Crowd Counting and Density Estimation (2017, background): “have demonstrated encouraging results [15, 20, 78].”
Learning To Count Objects in Images (2010, methodology): “Finally, counting by segmentation methods [10, 28] can be regarded as hybrids of counting-by-detection and counting-by-regression approaches.”
Bayesian Loss for Crowd Count Estimation With Point Supervision (2019, methodology): “Former methods [5, 35, 7] in this category rely on hand-crafted features, such as SIFT, LBP etc ., and then learn a regression model.”
Deep Count: Fruit Counting Based on Deep Simulated Learning (2017, unknown): “.”
Single-Image Crowd Counting via Multi-Column Convolutional Neural Network (2016, methodology): “The main steps of this kind of method are: 1) segmenting the foreground; 2) extracting various features from the foreground, such as area of crowd mask [4, 7, 27, 23], edge count [4, 7, 27, 25], or texture features [22, 7]; 3) utilizing a regression function to estimate the crowd count.”
Multi-source Multi-scale Counting in Extremely Dense Crowd Images (2013, background): “Such a function can be global [4, 6, 11, 21] where a single function’s parameters are learned for the entire image or video.”
Interactive Object Counting (2014, background): “The columns correspond to four training/testing splits of the data (‘maximal’,‘downscale’,‘upscale’,‘minimal’) proposed in [17].”
Crowd Counting and Density Estimation by Trellis Encoder-Decoder Networks (2019, background): “The successes of regression-based methods [3, 34, 10, 51, 15] can thus be attributed to their ability of circumventing explicit detection and directly mapping the input images to scalar values.”
Learning to Detect Partially Overlapping Instances (2013, methodology): “We follow the protocol from [22] and split the data into four groups in order to assess accuracy, scalability and practicality.”
Towards Perspective-Free Object Counting with Deep Learning (2016, methodology): “We follow exactly the same experimental setup that is used in [5, 6, 26, 7].”
Feature Mining for Localised Crowd Counting (2012, background): “Global approaches [5, 7, 14, 18, 21] learn a single regression function between image features extracted globally from the entire image space and the total people count in that image.”
Generating High-Quality Crowd Density Maps Using Contextual Pyramid CNNs (2017, methodology): “To overcome the issues of occlusion and high background clutter, researchers attempted to count by regression where they learn a mapping between features extracted from local image patches to their counts [3, 27, 6].”
CNN-Based cascaded multi-task learning of high-level prior and density estimation for crowd counting (2017, background): “To overcome these issues, researchers attempted to count by regression where they learn a mapping between features extracted from local image patches to their counts [22, 6].”
Cross-scene crowd counting via deep convolutional neural networks (2015, methodology): “Following the experiment settings in [12] and [24], we split the data into four different training and test sets: 1) ‘maximal’ :training on frames 600:5:1400; 2) ‘downscale’: training on frames 1205:5:1600; 3) ‘upscale’ : training on frames 805:5:1100; 4) ‘minimal’ : training on frames 640:80:1360.”
Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowds (2018, background): “Crowd counting is active an area of research with works tackling the three aspects of the problem: counting-by-regression [23], [17], [12], [4], [28], density map estimation [17], [7], [29], [20], [30] and localization [18], [22].”
Scale Aggregation Network for Accurate and Efficient Crowd Counting (2018, methodology): “Researchers attempted regression-based methods to directly learn a mapping from the feature of image patches to the count in the region [18,19,20].”