Michael Jeffrey Jones

Mitsubishi Electric Research Laboratories · Spatial Analytics

We investigate algorithms for efficiently detecting anomalies in real-valued one-dimensional time series. Past work has shown that a simple brute force algorithm that uses as an anomaly score the Euclidean distance between nearest neighbors of subsequences from a testing time series and a training time series is one of the most effective anomaly de...

We present a real-time system for 3D head pose estimation and facial landmark localization using a commodity depth sensor. We introduce a novel triangular surface patch (TSP) descriptor, which encodes the shape of the 3D surface of the face within a triangular area. The proposed descriptor is viewpoint invariant, and it is robust to noise and to va...

A climate control unit is controlled by constructing background and foreground models of an environment from images acquired of the environment by a camera. The background model represents the environment when unoccupied, and there is one foreground model for each person in the environment. A 2D location of each person in the environment is determi...

Systems, methods, and media that: implement a boosted classifier having a plurality of weak hypotheses that produce a classification, each of the plurality of weak hypotheses having at least one weight; receive testing data; receive at least one piece of training data subsequently to receiving the testing data; calculate corrective terms for correc...

This paper proposes a person tracking framework using a scanning low-resolution thermal infrared (IR) sensor colocated with a wide-angle RGB camera. The low temporal and spatial resolution of the low-cost IR sensor make it unable to track moving people and prone to false detections of stationary people. Thus, IR-only tracking using only this sensor...

We study the problem of predicting travel times for links (road segments) using floating car data. We present four different methods for predicting travel times and discuss the differences in predicting on congested and uncongested roads. We show that estimates of the current travel time are mainly useful for prediction on links that get congested....

A method for detecting an object in a depth image includes determining a detection window covering a region in the depth image, wherein a location of the detection window is based on a location of a candidate pixel in the depth image, wherein a size of the detection window is based on a depth value of the candidate pixel and a size of the object. A...

An ideal approach to the problem of pose-invariant face recognition would handle continuous pose variations, would not be database specific, and would achieve high accuracy without any manual intervention. Most of the existing approaches fail to match one or more of these goals. In this paper, we present a fully automatic system for pose-invariant...

We present a novel approach to pose-invariant face recognition that handles continuous pose variations, is not database-specific, and achieves high accuracy without any manual intervention. Our method uses multi dimensional Gaussian process regression to learn a nonlinear mapping function from the 2D shapes of faces at any non-frontal pose to the c...

In this paper, we present a novel framework to address the confounding effects of illumination variation in face recognition. By augmenting the gallery set with realisti- cally relit images, we enhance recognition performance in a classier -independent way. We describe a novel method for single-image relighting, Morphable Reectance Fields (MoRF), w...

The purpose of this work is to speed-up mar-gin based learning algorithms by deriving a stopping rule which stops the partial compu-tation of the margin when the full compu-tation is likely to have the same outcome. This is achieved by a novel merger of se-quential analysis and margin based learning. Early stopping may introduce decision errors, wh...

We present a new online boosting algorithm for updating the weights of a boosted classifier, which yields a closer approximation to the edges found by Freund and Schapire's AdaBoost algorithm than previous online boosting algorithms. We contribute a new way of deriving the online algorithm that ties together previous online boosting work. The onlin...

Face recognition has become a very active research field. Despite the existence of commercial face recognition systems, there are still important challenges for further research. The problems of variable lighting, pose, facial expression, aging, and inaccurate alignment continue to cause larger than desired error rates. This paper discusses the cur...

A scanning window type pedestrian detector is presented that uses both appearance and motion information to find walking people in surveillance video. We extend the work of Viola, Jones and Snow (2005) to use many more frames as input to the detector thus allowing a much more detailed analysis of motion. The resulting detector is about an order of...

We present a new online boosting algorithm for adapting the weights of a boosted classifier, which yields a closer approximation to Freund and Schapire's AdaBoost algorithm than previous online boosting algorithms. We also contribute a new way of deriving the online algorithm that ties together previous online boosting work. We assume that the weak...

MERL’s recognizer has two major components (Figure 1). The first is an aligner whose input is any arbitrary image and whose output is a cropped and rectified face if the image contains a face. The second is a comparator whose input is two aligned faces and whose output is a similarity score. The comparator is described in detail in [3]. The aligner...

This paper explores the use of thresholded hyperplanes as the building blocks of a classifier for face detection. We are motivated by the work of Viola and Jones [10] who used Haar-like wavelet features as their weak classifiers in the AdaBoost learning algorithm. These weak classifiers were chosen for their speed. We explore how much may be gained...

Biometrics has become more and more important in security applications. In comparison with many other bio- metric features, iris recognition has very high recognition accuracy. Successful iris recognition depends largely on correct iris localization, however, the performance of current techniques for iris localization still leaves room for improvem...

Biometrics is increasingly important in security applications. Iris recognition provides the great-est accuracy among known biometrics. The accuracy of iris recognition is, for example, much greater than face recognition and fingerprint recognition. However, it is not trivial to capture iris images in practice, and usually the users need to adjust...

This paper describes a pedestrian detection system that integrates image intensity information with motion information. We use a detection style algorithm that scans a detector over two consecutive frames of a video sequence. The detector is trained (using AdaBoost) to take advantage of both motion and appearance information to detect a walking per...

This paper describes a machine learning approach for visual object detection which is capable of processing images extremely rapidly and achieving high detection rates. This work is distinguished by three key contributions. The first is the introduction of a new image representation called the Integral Image which allows the features used by our de...

This paper describes a face detection framework that is capable of processing images extremely rapidly while achieving high detection rates. There are three key contributions. The first is the introduction of a new image representation called the Integral Image which allows the features used by our detector to be computed very quickly. The second i...

This paper describes a face detection framework that is capable of processing images extremely rapidly while achieving high detection rates. There are three key contributions. The first is the introduction of a new image representation called the “Integral Image” which allows the features used by our detector to be computed very quickly. The second...

This paper describes a machine learning approach for visual object detection which is capable of processing images extremely rapidly and achieving high detection rates. This work is distinguished by three key contributions. The first is the introduction of a new image representation called the "Integral Image" which allows the features used by our...

The existence of large image datasets such as the set of photos on the World Wide Web make it possible to build powerful generic models for low-level image attributes like color using simple histogram learning techniques. We describe the construction of color models for skin and non-skin classes from a dataset of nearly 1 billion labelled pixels. T...

This paper extends the face detection framework proposed by Viola and Jones 2001 to handle profile views and rotated faces. As in the work of Rowley et al. 1998, and Schneiderman et al. 2000, we build different detectors for different views of the face. A decision tree is then trained to determine the viewpoint class (such as right profile or rotat...

This paper extends the face detection framework proposed by Viola and Jones 2001 to handle profile views and rotated faces. As in the work of Rowley et al 1998. and Schneiderman et al. 2000, we build different detectors for different views of the face

This paper describes a pedestrian detection system that integrates image intensity information with motion information. We use a detection style algorithm that scans a detector over two consecutive frames of a video sequence. The detector is trained (using AdaBoost) to take advantage of both motion and appearance information to detect a walking per...

This paper presents a new method for face recognition which learns a face similarity measure from example image pairs. A set of computationally efficient "rectangle" features are described which act on pairs of input images. The features compare regions within the input images at different locations, scales, and orientations. The AdaBoost algorithm...

This paper describes a pedestrian detection system that integrates image intensity information with motion information. We use a detection style algorithm that scans a detector over two consecutive frames of a video sequence. The detector is trained (using AdaBoost) to take advantage of both motion and appearance information to detect a walking per...

In this paper, we address the problem of recovering 3-D models from sequences of partly calibrated images with unknown correspondence. To that end, we integrate tracking, structure from motion with geometric constraints (specifically in the form of linear class models) in a single framework. The key to making the proposed approach work is the use o...

This paper describes a visual object detection framework that is capable of processing images extremely rapidly while achieving high detection rates. There are three key contributions. The first is the introduction of a new image representation called the "Integral Image" which allows the features used by our detector to be computed very quickly. T...

This paper develops a new approach for extremely fast detection in domains where the distribution of positive and negative examples is highly skewed (e.g. face detection or database retrieval). In such domains a cascade of simple classifiers each trained to achieve high detection rates and modest false positive rates can yield a final detector with...

We propose a mathematical formulation for the notion of optimal projective cluster, starting from natural requirements on the density of points in subspaces. This allows us to develop a Monte Carlo algorithm for iteratively computing projective clusters. We prove that the computed clusters are good with high probability. We implemented a modified v...

We had previously shown that regularization principles lead to approximation schemes which are equivalent to networks with one layer of hidden units, called Regularization Networks. In particular we had discussed how standard smoothness functionals lead to a subclass of regularization networks, the well-known Radial Basis Functions approximation sc...

Flexible models of object classes, based on linear combinations of prototypical images, are capable of matching novel images of the same class and havebeenshown to be a powerful tool to solveseveral fundamental vision tasks such as recognition, synthesis and correspondence. The key problem in creating a specific flexible model is the computation of...

We describe a flexible model similar to (Vetter and Poggio, 1993, 1995 and Jones and Poggio, 1995) for representing images of objects of a certain class, known a priori, suchas faces, and introduce a new algorithm for matching it to a novel image and thereby perform image analysis. The flexible model is learned from example images (called prototype...

This paper describes a machine learning approach for visual object detection which is capable of processing images extremely rapidly and achieving high detection rates. This work is distinguished by three key contributions. The first is the introduction of a new image representation called the "integral image" which allows the features used by our...

This paper describes a visual object detection framework that is capable of processing images extremely rapidly while achieving high detection rates. There are three key contributions. The first is the introduction of a new image representation called the "Integral Image" which allows the features used by our detector to be computed very quickly. T...

We have constructed a frontal face detection system which achieves detection and false positive rates which are equivalent to the best published results [7, 5, 6, 4, 1]. This face detection system is most clearly distinguished from previous approaches in its ability to detect faces extremely rapidly. Operating on 384 by 288 pixel images, faces are...

This paper develops a new approach for extremely fast detection in domains where the distribution of positive and negative examples is highly skewed (e.g. face detection or database retrieval). In such domains a cascade of simple classifiers each trained to achieve high detection rates and modest false positive rates can yield a final detector with...

This paper describes a visual object detection framework that is capable of processing images extremely rapidly while achieving high detection rates. There are three key contributions. The first is the introduction of a new image representation called the "Integral Image" which allows the features used by our detector to be computed very quickly. T...

This paper presents preliminary work on a novel technique for gaze estimation from a single image. The goal is to provide rough estimates of where a person is looking at a monitor. Many applications for human-computer interaction are possible for such a technique. Our approach uses the morphable model framework of Jones and Poggio [4] to model a re...

The existence of large image datasets such as photos on the World Wide Web make it possible to build powerful generic models for low-level image attributes like color using simple histogram learning techniques. We describe the construction of color models for skin and non-skin classes from a dataset of nearly 1 billion labeled pixels. These classes...

This paper presents an approach to object detection which is based on recent work in statistical models for texture synthesis and recognition. Our method follows the texture recognition work of De Bonet and Viola (1998). We use feature vectors which capture the joint occurrence of local features at multiple resolutions. The distribution of feature...

We describe a flexible model similar to (Vetter and Poggio, 1993, 1995 and Jones and Poggio, 1995) for representing images of objects of a certain class, known a priori, such as faces, and introduce a new algorithm for matching it to a novel image and thereby perform image analysis. The flexible model is learned from example images (called prototyp...

We had previously shown that regularization principles lead to approximation schemes which are equivalent to networks with one layer of hidden units, called Regularization Networks. In particular, standard smoothness functionals lead to a subclass of regularization networks, the well known Radial Basis Functions approximation schemes. This paper sh...

We describe a flexible model for representing images of objects of a certain class, known a priori, such as faces, and introduce a new algorithm for matching it to a novel image and thereby perform image analysis. The flexible model, known as a multidimensional morphable model, is learned from example images of objects of a class. In this paper we...

The paper presents preliminary work on a novel technique for gaze estimation from a single image. The goal is to provide rough estimates of where a person is looking at a monitor. Many applications for human-computer interaction are possible for such a technique. The approach uses the morphable model framework of Jones and Poggio (1998) to model a...

We describe a flexible model for representing images of objects of a certain class, known a priori, such as faces, and introduce a new algorithm for matching it to a novel image and thereby performing image analysis. We call this model a multidimensional morphable model or just a, morphable model. The morphable model is learned from example images...

Perceptual tasks such as edge detection, image segmentation, lightness computation and estimation of three-dimensional structure are considered to be low-level or mid-level vision problems and are traditionally approached in a bottom-up, generic and hard-wired way. An alternative to this would be to take a top-down, object-class-specific and exampl...

This paper presents a new technique for modelling object classes (such as faces) and matching the model to novel images from the object class. The technique can be used for a variety of image analysis applications including face recognition, object verification and facial expression analysis. The model, called a hierarchical morphable model, is “le...

Flexible models of object classes, based on linear combinations of prototypical images, are capable of matching novel images of the same class and have been shown to be a powerful tool to solve several fundamental vision tasks such as recognition, synthesis and correspondence. The key problem in creating a specific flexible model is the computation...

. Flexible models of object classes, based on linear combinations of prototypical images, are capable of matching novel images of the same class and have been shown to be a powerful tool to solve several fundamental vision tasks such as recognition, synthesis and correspondence. The key problem in creating a specific flexible model is the computati...

We describe a technique for finding pixelwise correspondences between two images by using models of objects of the same class to guide the search. The object models are "learned" from example images (also called prototypes) of an object class. The models consist of a linear combination of prototypes. The flow fields giving pixelwise correspondences...

This thesis explores how recurrent neural networks can be exploited for learning certain high-dimensional mappings. Recurrent networks are shown to be as powerful as Turing machines in terms of the class of functions they can compute. Given this computational power, a natural question to ask is how recurrent networks can be used to simplify the pro...

We had previously shown that regularization principles lead to approximation schemes which are equivalent to networks with one layer of hidden units, called regularization networks. We summarize some recent results (Girosi, Jones and Poggio, 1993) that show that regularization networks encompass a much broader range of approximation schemes, includ...

This paper presents a new method for face recognition which learns a face similarity measure from example image pairs. A set of computationally efficient "rectangle" features are described which act on pairs of input images. The features compare regions within the input images at different locations, scales, and orientations. The AdaBoost algorithm...

The purpose of this work is to lower the average num-ber of features that are evaluated by an online algorithm. This is achieved by merging Sequential Analysis and Online Learning. Many online algorithms use the example's mar-gin to decide whether the model should be updated. Usu-ally, the algorithm's model is updated when the margin is smaller tha...