A High Accuracy Pedestrian Detection System Combining a Cascade AdaBoost Detector and Random Vector Functional-Link Net

Abstract

In pedestrian detection methods, their high accuracy detection rates are always obtained at the cost of a large amount of false pedestrians. In order to overcome this problem, the authors propose an accurate pedestrian detection system based on two machine learning methods: cascade AdaBoost detector and random vector functional-link net. During the offline training phase, the parameters of a cascade AdaBoost detector and random vector functional-link net are trained by standard dataset. These candidates, extracted by the strategy of a multiscale sliding window, are normalized to be standard scale and verified by the cascade AdaBoost detector and random vector functional-link net on the online phase. Only those candidates with high confidence can pass the validation. The proposed system is more accurate than other single machine learning algorithms with fewer false pedestrians, which has been confirmed in simulation experiment on four datasets.

Intelligent Pedestrian Detection using Optical Flow and HOG

“…In order to reduce this false detection rate Z. Wang et al [34] presented a two stage machine learning algorithms based approach for efficient and accurate pedestrian detection. This approach is based on highly efficient combination of cascade AdaBoost detector and vector function link net derived from machine learning domain.…”

Pyramidal Part-Based Model for Partial Occlusion Handling in Pedestrian Classification

“…e functional-link net [39] used the cascade AdaBoost detector and random vector functional-link net to enhance the detection accuracy and reduce the number of false positive rates for pedestrian detection. For the fast multiscale object detection, the multiscale CNN (MS-CNN) [47] was proposed to produce accurate object proposals on the detection network with the use of feature upsampling.…”

“…In the classification method, the well-known techniques such as support vector machine (linear and latent) [4,5,10,12,25,38], AdaBoost [7,9,34,39], neural network [2,3], random forest [3], and cascade [7,17,27] had been widely used. Among them, the linear SVM is one of the useful techniques that can compute faster than a latent SVM in terms of performance and efficiency.…”

An improved cuckoo search based extreme learning machine for medical data classification

“…extreme learning machine (ELM) [11]. Researchers have shown that ELM owes its origin to random vector functional link (RVLF) [28][29][30][31][32][33][34][35]. However Huang [36] has shown its newness in Ref.…”

“…Literature survey reveals that ELM has been extensively used in many applications [37][38][39][40][41][42][43]. Although several variants of extreme learning machines [28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43] are now available for multiclass classification there remains several problems like the optimal choice of network size requiring a large number of hidden nodes for better generalization and choice of activation functions. Besides the randomness of ELM causes an additional uncertainty in regression and classification problems with regards to universal approximation and learning.…”