Speed up deep neural network based pedestrian detection by sharing features across multi-scale models
NeurocomputingPublished 24 December 2015Open access
Xiaoheng Jiang, Yanwei Pang, Xuelong Li, Jing Pan
Citations55
SJR quartileQ1
SJR score1.47
SNIP1.94
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
This paper proposes to share features across a group of DNNs that correspond to pedestrian models of different sizes that can detect pedestrians of several different scales on one single layer of an image pyramid to improve detection efficiency.
Abstract
Deep neural networks (DNNs) have now demonstrated state-of-the-art detection performance on pedestrian datasets. However, because of their high computational complexity, detection efficiency is still a frustrating problem even with the help of Graphics Processing Units (GPUs). To improve detection efficiency, this paper proposes to share features across a group of DNNs that correspond to pedestrian models of different sizes. By sharing features, the computational burden for extracting features from an image pyramid can be significantly reduced. Simultaneously, we can detect pedestrians of several different scales on one single layer of an image pyramid. Furthermore, the improvement of detection efficiency is achieved with negligible loss of detection accuracy. Experimental results demonstrate the robustness and efficiency of the proposed algorithm. (C) 2015 The Authors. Published by Elsevier B.V.
Keywords
Computer Science
Communications of the ACMImageNet classification with deep convolutional neural networks
75,709 Citations2017Alex Krizhevsky, Ilya Sutskever +1 more
A large, deep convolutional neural network was trained to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes and employed a recently developed regularization method called "dropout" that proved to be very effective.
arXiv (Cornell University)Very Deep Convolutional Networks for Large-Scale Image Recognition
75,540 Citations2014Karen Simonyan, Andrew Zisserman
This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.
Histograms of Oriented Gradients for Human Detection
31,896 Citations2005Navneet Dalal, Bill Triggs
It is shown experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection, and the influence of each stage of the computation on performance is studied.
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
31,752 Citations2014Ross Girshick, Jeff Donahue +2 more
This paper proposes a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%.
Lecture notes in computer scienceVisualizing and Understanding Convolutional Networks
15,444 Citations2014Matthew D. Zeiler, Rob Fergus
A novel visualization technique is introduced that gives insight into the function of intermediate feature layers and the operation of the classifier in large Convolutional Network models, used in a diagnostic role to find model architectures that outperform Krizhevsky et al on the ImageNet classification benchmark.
Caffe
11,194 Citations2014Yangqing Jia, Evan Shelhamer +6 more
Caffe provides multimedia scientists and practitioners with a clean and modifiable framework for state-of-the-art deep learning algorithms and a collection of reference models for training and deploying general-purpose convolutional neural networks and other deep models efficiently on commodity architectures.
IEEE Transactions on Pattern Analysis and Machine IntelligenceObject Detection with Discriminatively Trained Part-Based Models
9,976 Citations2009Pedro F. Felzenszwalb, Ross Girshick +2 more
An object detection system based on mixtures of multiscale deformable part models that is able to represent highly variable object classes and achieves state-of-the-art results in the PASCAL object detection challenges is described.
Large-Scale Video Classification with Convolutional Neural Networks
6,282 Citations2014Andrej Karpathy, George Toderici +4 more
This work studies multiple approaches for extending the connectivity of a CNN in time domain to take advantage of local spatio-temporal information and suggests a multiresolution, foveated architecture as a promising way of speeding up the training.
International Journal of Computer VisionSelective Search for Object Recognition
6,118 Citations2013Jasper Uijlings, Koen E. A. van de Sande +2 more
This paper introduces selective search which combines the strength of both an exhaustive search and segmentation, and shows that its selective search enables the use of the powerful Bag-of-Words model for recognition.
arXiv (Cornell University)Two-Stream Convolutional Networks for Action Recognition in Videos
5,369 Citations2014Karen Simonyan, Andrew Zisserman
This work proposes a two-stream ConvNet architecture which incorporates spatial and temporal networks and demonstrates that a ConvNet trained on multi-frame dense optical flow is able to achieve very good performance in spite of limited training data.
CNN Features Off-the-Shelf: An Astounding Baseline for Recognition
4,341 Citations2014Ali Sharif Razavian, Hossein Azizpour +2 more
A series of experiments conducted for different recognition tasks using the publicly available code and model of the OverFeat network which was trained to perform object classification on ILSVRC13 suggest that features obtained from deep learning with convolutional nets should be the primary candidate in most visual recognition tasks.
IEEE Transactions on Pattern Analysis and Machine IntelligencePedestrian Detection: An Evaluation of the State of the Art
3,221 Citations2011Piotr Dollár, Christian Wojek +2 more
An extensive evaluation of the state of the art in a unified framework of monocular pedestrian detection using sixteen pretrained state-of-the-art detectors across six data sets and proposes a refined per-frame evaluation methodology.
Lecture notes in computer scienceSpatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
3,113 Citations2014Kaiming He, Xiangyu Zhang +2 more
This work equips the networks with another pooling strategy, “spatial pyramid pooling”, to eliminate the above requirement, and develops a new network structure, called SPP-net, which can generate a fixed-length representation regardless of image size/scale.
Lecture notes in computer scienceEdge Boxes: Locating Object Proposals from Edges
2,766 Citations2014C. Lawrence Zitnick, Piotr Dollár
A novel method for generating object bounding box proposals using edges is proposed, showing results that are significantly more accurate than the current state-of-the-art while being faster to compute.
arXiv (Cornell University)OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
1,332 Citations2013Pierre Sermanet, David Eigen +4 more
Pedestrian Detection with Unsupervised Multi-stage Feature Learning
749 Citations2013Pierre Sermanet, Koray Kavukcuoglu +2 more
This work reports state-of-the-art and competitive results on all major pedestrian datasets with a convolutional network model that uses a few new twists, such as multi-stage features, connections that skip layers to integrate global shape information with local distinctive motif information.
Learning to generate chairs with convolutional neural networks
683 Citations2015Alexey Dosovitskiy, Jost Tobias Springenberg +1 more
This work trains a generative convolutional neural network which is able to generate images of objects given object type, viewpoint, and color and shows that the network can be used to find correspondences between different chairs from the dataset, outperforming existing approaches on this task.
Joint Deep Learning for Pedestrian Detection
681 Citations2013Wanli Ouyang, Xiaogang Wang
This paper forms these four important components in pedestrian detection into a joint deep learning framework and proposes a new deep network architecture that achieves a 9% reduction in the average miss rate compared with the current best-performing pedestrian detection approaches on the largest Caltech benchmark dataset.
Lecture notes in computer scienceTen Years of Pedestrian Detection, What Have We Learned?
548 Citations2015Rodrigo Benenson, Mohamed Omran +2 more
This work analyzes the remarkable progress of the last decade by dis- cussing the main ideas explored in the 40+ detectors currently present in the Caltech pedestrian detection benchmark to find a new decision forest detector.
New features and insights for pedestrian detection
458 Citations2010Stefan Walk, Nikodem Majer +2 more
This work shows that motion features derived from optic flow yield substantial improvements on image sequences, if implemented correctly — even in the case of low-quality video and consequently degraded flow fields, and introduces a new feature, self-similarity on color channels, which consistently improves detection performance across different datasets.
DeepID-Net: Deformable deep convolutional neural networks for object detection
438 Citations2015Wanli Ouyang, Xiaogang Wang +9 more
The proposed approach improves the mean averaged precision obtained by RCNN, which was the state-of-the-art, from 31% to 50.3% on the ILSVRC2014 detection test set.
Lecture notes in computer scienceAnalyzing the Performance of Multilayer Neural Networks for Object Recognition
393 Citations2014Pulkit Agrawal, Ross Girshick +1 more
This paper experimentally probes several aspects of CNN feature learning in an attempt to help practitioners gain useful, evidence-backed intuitions about how to apply CNNs to computer vision problems.
A discriminative deep model for pedestrian detection with occlusion handling
297 Citations2012Wanli Ouyang, Xiaogang Wang
A probabilistic pedestrian detection framework using a deformable part-based model to obtain the scores of part detectors and the visibilities of parts are modeled as hidden variables and a discriminative deep model is used for learning the visibility relationship among overlapping parts at multiple layers.
Fast image scanning with deep max-pooling convolutional neural networks
296 Citations2013Alessandro Giusti, Dan Cireşan +3 more
This work shows how dynamic programming can speedup the process by orders of magnitude, even when max-pooling layers are present.
Taking a deeper look at pedestrians
291 Citations2015Jan Hosang, Mohamed Omran +2 more
This paper analyses small and big convnets, their architectural choices, parameters, and the influence of different training data, including pretraining on surrogate tasks, and presents the best convnet detectors on the Caltech and KITTI dataset.
arXiv (Cornell University)Descriptor Matching with Convolutional Neural Networks: a Comparison to SIFT
247 Citations2014Philipp Fischer, Alexey Dosovitskiy +1 more
This paper compares features from various layers of convolutional neural nets to standard SIFT descriptors and Surprisingly, convolutionAL neural networks clearly outperform SIFT on descriptor matching.
Switchable Deep Network for Pedestrian Detection
235 Citations2014Ping Luo, Yonglong Tian +2 more
A new generative algorithm is devised to effectively pretrain the SDN and then fine-tune it with back-propagation to achieve the state-of-the-art detection performance.
NeurocomputingAU-inspired Deep Networks for Facial Expression Feature Learning
232 Citations2015Mengyi Liu, Shaoxin Li +2 more
A deep architecture, AU-inspired Deep Networks (AUDN), inspired by the psychological theory that expressions can be decomposed into multiple facial Action Units (AUs), which can achieve state-of-the-art results on all the databases, and validates the effectiveness of AUDN in both lab-controlled and wild environments.
Multi-stage Contextual Deep Learning for Pedestrian Detection
153 Citations2013Xingyu Zeng, Wanli Ouyang +1 more
A new deep model is proposed that can jointly train multi-stage classifiers through several stages of back propagation and keeps the score map output by a classifier within a local region and uses it as contextual information to support the decision at the next stage.
Single-Pedestrian Detection Aided by Multi-pedestrian Detection
146 Citations2013Wanli Ouyang, Xiaogang Wang
A probabilistic framework is proposed to model the relationship between the configurations estimated by single-and multi-Pedestrian detectors, and to refine the single-pedestrian detection result with multi-pedESTrian detection, which can integrate with any single- Pedestrian detector without significantly increasing the computation load.
arXiv (Cornell University)DeepID-Net: multi-stage and deformable deep convolutional neural networks for object detection
134 Citations2014Wanli Ouyang, Ping Luo +13 more
A set of models with large diversity are obtained, which significantly improves the effectiveness of modeling averaging, and improves the mean averaged precision obtained by RCNN, which is the state of the art of object detection, from31% to 45%.
IEEE Transactions on CyberneticsDistributed Object Detection With Linear SVMs
77 Citations2014Yanwei Pang, Kun Zhang +2 more
NeurocomputingAn improved radial basis function neural network for object image retrieval
62 Citations2015Gholam Ali Montazer, Davar Giveki
A new learning method for RBFNNs is presented and a new method for object image retrieval is proposed, using the improved radial basis function neural network, which achieves high performance.
Signal ProcessingFlexible sliding windows with adaptive pixel strides
14 Citations2014Xiaoheng Jiang, Yanwei Pang +2 more
This paper proposes to scan the image using adaptive instead of fixed (unchanged) strides, which results in the reduction of total computation cost without loss of detection accuracy, and applies sparsely sampling on background and densely sampling on potential object regions.
Signal ProcessingEfficient object detection by prediction in 3D space
11 Citations2014Yanwei Pang, Xiaoheng Jiang +2 more
This paper proposes to explore both the scaling capacity and translation capacity of object detector to accelerate detection speed, without loss of detection accuracy, and presents an interlaced scanning method called diamond grid scanning which can reduce redundant evaluation.
NeurocomputingGoal-oriented behavior sequence generation based on semantic commands using multiple timescales recurrent neural network with initial state correction
9 Citations2013Sungmoon Jeong, Yunjung Park +3 more
Experimental results show that the proposed scheme can successfully generate suitable behavior timing, for a robot to autonomously achieve the tasks involving multiple objects, such as searching, approaching and hitting the target object using its arm.