Szu-Hao Huang

Learning-Based Vertebra Detection and Iterative Normalized-Cut Segmentation for Spinal MRI

Automatic extraction of vertebra regions from a spinal magnetic resonance (MR) image is normally required as the first step to an intelligent spinal MR image diagnosis system. In this work, we develop a fully automatic vertebra detection and segmentation system, which consists of three stages; namely, AdaBoost-based vertebra detection, detection refinement via robust curve fitting, and vertebra segmentation by an iterative normalized cut algorithm. In order to produce an efficient and effective vertebra detector, a statistical learning approach based on an improved AdaBoost algorithm is proposed. A robust estimation procedure is applied on the detected vertebra locations to fit a spine curve, thus refining the above vertebra detection results. This refinement process involves removing the false detections and recovering the miss-detected vertebrae. Finally, an iterative normalized-cut segmentation algorithm is proposed to segment the precise vertebra regions from the detected vertebra locations. In our implementation, the proposed AdaBoost-based detector is trained from 22 spinal MR volume images. The experimental results show that the proposed vertebra detection and segmentation system can achieve nearly 98% vertebra detection rate and 96% segmentation accuracy on a variety of testing spinal MR images. Our experiments also show the vertebra detection and segmentation accuracies by using the proposed algorithm are superior to those of the previous representative methods. The proposed vertebra detection and segmentation system is proved to be robust and accurate so that it can be used for advanced research and application on spinal MR images.

Improved AdaBoost-based image retrieval with relevance feedback via paired feature learning

Boost learning algorithm, such as AdaBoost, has been widely used in a variety of applications in multimedia and computer vision. Relevance feedback-based image retrieval has been formulated as a classification problem with a small number of training samples. Several machine learning techniques have been applied to this problem recently. In this paper, we propose a novel paired feature AdaBoost learning system for relevance feedback-based image retrieval. To facilitate density estimation in our feature learning method, we propose an ID3-like balance tree quantization method to preserve most discriminative information. By using paired feature combination, we map all training samples obtained in the relevance feedback process onto paired feature spaces and employ the AdaBoost algorithm to select a few feature pairs with best discrimination capabilities in the corresponding paired feature spaces. In the AdaBoost algorithm, we employ Bayesian classification to replace the traditional binary weak classifiers to enhance their classification power, thus producing a stronger classifier. Experimental results on content-based image retrieval (CBIR) show superior performance of the proposed system compared to some previous methods.

Detecting Faces from Color Video by Using Paired Wavelet Features

Detecting human face regions in color video is normally required for further processing in many practical applications. In this paper, we propose a learning-based algorithm that determines the most discriminative pairs of Haar wavelet coefficients of color images for face detection. To select the most discriminative features from the vast amount (1,492,128) of possible pairs of three-channel color wavelet coefficients, we employ two procedures to accomplish this task. At first, we choose a subset of effective candidate pairs of wavelet coefficients based on the Kullback Leibler (KL) distance between the conditional joint distributions of the face and non-face training data. Then, the adaboost algorithm is employed to incrementally select a set of complementary pairs of wavelet coefficients and determine the best combination of weak classifiers that are based on the joint conditional probabilities of these selected coefficient pairs for face detection. By applying Kalman filter to predict and update the face region in a video, we extending the face detection from a single image to a video sequence. In contrast to the previous face detection works, the proposed algorithm is based on finding the discriminative features of joint wavelet coefficients computed from all three channels of color images in an integrated learning framework. We experimentally show that the proposed algorithm can achieve high accuracy and fast speed for detecting faces from color video.

People localization in a camera network combining background subtraction and scene-aware human detection

In a network of cameras, people localization is an important issue. Traditional methods utilize camera calibration and combine results of background subtraction in different views to locate people in the three dimensional space. Previous methods usually solve the localization problem iteratively based on background subtraction results, and high-level image information is neglected. In order to fully exploit the image information, we suggest incorporating human detection into multi-camera video surveillance. We develop a novel method combining human detection and background subtraction for multi-camera human localization by using convex optimization. This convex optimization problem is independent of the image size. In fact, the problem size only depends on the number of interested locations in ground plane. Experimental results show this combination performs better than background subtraction-based methods and demonstrate the advantage of combining these two types of complementary information.

A statistical learning appproach to vertebra detection and segmentation from spinal MRI

Automatically extracting vertebra regions from a spinal magnetic resonance image is normally required as the first step to an intelligent spinal MR image diagnosis system. In this work, we develop a fully automatic vertebra detection and segmentation method. Our system consists of three stages; namely, AdaBoost-based vertebra detection, detection refinement via robust curve fitting, and vertebra segmentation by an iterative normalized cut algorithm. We proposed an efficient and effective vertebra detector, which is trained by the improved AdaBoost algorithm, to locate the initial vertebra positions. Then, a robust estimation procedure is applied to fit all the vertebrae as a polynomial spinal curve to refine the vertebra detection results. Finally, an iterative segmentation algorithm based on normalized-cut energy minimization is applied to extract the precise vertebra regions from the detected windows. The experimental results show our system can achieve high accuracy on a number of testing 3D spinal MRI data sets.

A learning-based contrarian trading strategy via a dual-classifier model

Behavioral finance is a relatively new and developing research field which adopts cognitive psychology and emotional bias to explain the inefficient market phenomenon and some irrational trading decisions. Unlike the experts in this field who tried to reason the price anomaly and applied empirical evidence in many different financial markets, we employ the advanced binary classification algorithms, such as AdaBoost and support vector machines, to precisely model the overreaction and strengthen the portfolio compositions of the contrarian trading strategies. The novelty of this article is to discover the financial time-series patterns through a high-dimensional and nonlinear model which is constructed by integrated knowledge of finance and machine learning techniques. We propose a dual-classifier learning framework to select candidate stocks from the past results of original contrarian trading strategies based on the defined learning targets. Three different feature extraction methods, including wavelet transformation, historical return distribution, and various technical indicators, are employed to represent these learning samples in a 381-dimensional financial time-series feature space. Finally, we construct the classifier models with four different learning kernels and prove that the proposed methods could improve the returns dramatically, such as the 3-year return that improved from 26.79% to 53.75%. The experiments also demonstrate significantly higher portfolio selection accuracy, improved from 57.47% to 66.41%, than the original contrarian trading strategy. To sum up, all these experiments show that the proposed method could be extended to an effective trading system in the historical stock prices of the leading U.S. companies of S&P 100 index.

Real-time video surveillance based on combining foreground extraction and human detection

In this paper, we present an adaptive foreground object extraction algorithm for real-time video surveillance, in conjunction with a human detection technique applied in the extracted foreground regions by using AdaBoost learning algorithm and Histograms of Oriented Gradient (HOG) descriptors. Furthermore, a RANSAC-based temporal tracking algorithm is also applied to refine and trace the detected human windows in order to increase the detection accuracy and reduce the false alarm rate. The traditional background subtraction technique usually cannot work well for situations with lighting variations in the scene. The proposed algorithm employs a two-stage foreground/background classification procedure to perform background subtraction and remove the undesirable subtraction results due to shadow, automatic white balance, and sudden illumination change. Experimental results on some real surveillance video are shown to demonstrate the good performance of the proposed adaptive foreground extraction algorithm under a variety of different environments with lighting variations and human detection system.

Intelligent home video management system

An integrated intelligent home video management system was proposed in this paper. Five different types of multimedia representative features were computed as the basis of the home video management system. With the aid of some machine learning techniques, such as SVM, neural network, adaboost algorithm, and the K-means clustering algorithm, we develop six main applications based on this system. These applications include detection of abnormal camera operation, shot boundary detection, fast-pan detection, face shot identification, key-frame extraction, and variable length video abstraction. Our system can help inexpert digital camcorder users manage their home video effectively in a smart way.

Improved adaboost-based image retrieval with relevance feedback via paired feature learning

In this paper, we propose a novel paired feature learning system for relevance feedback based image retrieval. To facilitate density estimation in our feature learning system, we employ an ID3-like balance tree quantization method to preserve most discriminative information. In addition, we map all training samples in the relevance feedback onto paired feature spaces to enhance the discrimination power of feature representation. Furthermore, we replace the traditional binary classifiers in the AdaBoost learning algorithm by Bayesian weak classifiers to improve its accuracy, thus producing stronger classifiers. Experimental results on content-based image retrieval show improvement of each step in the proposed learning system.

Real-time face detection in color video

In this paper, we propose a novel and fast face detection algorithm for detecting face in color video sequences. This algorithm can be integrated into a real-time surveillance or a video retrieval in the design of the algorithm. A set of multiresolution Haar wavelet coefficients pairs is selected by the proposed learning algorithm to determine if a particular region is a face. We apply an ID3-like balanced decision tree for the wavelet coefficients quantization, to reduce the quantization error. For each pair of quantized features, we estimate the associated conditional joint probability density function from a large set of face and non-face training data. Then, we compute the Kullback Leibler (KL) distance to measure the discrimination between the face and non-face conditional density functions for each feature pair. The feature pairs with larger KL-distance are selected as the feature candidates. It is an effective feature dimension reduction method and helps to speedup the Adaboost training algorithm when considering the spatial relationship between all coefficient pairs. Aided by an automatic skin color judgment method and a Gaussian face location model both in temporal and spatial domain, the experiments show that the proposed algorithm runs faster than 4 times the video rate with good detection accuracy.