Gary K. L. Tam

Registration of 3D Point Clouds and Meshes: A Survey from Rigid to Nonrigid

Three-dimensional surface registration transforms multiple three-dimensional data sets into the same coordinate system so as to align overlapping components of these sets. Recent surveys have covered different aspects of either rigid or nonrigid registration, but seldom discuss them as a whole. Our study serves two purposes: 1) To give a comprehensive survey of both types of registration, focusing on three-dimensional point clouds and meshes and 2) to provide a better understanding of registration from the perspective of data fitting. Registration is closely related to data fitting in which it comprises three core interwoven components: model selection, correspondences and constraints, and optimization. Study of these components 1) provides a basis for comparison of the novelties of different techniques, 2) reveals the similarity of rigid and nonrigid registration in terms of problem representations, and 3) shows how overfitting arises in nonrigid registration and the reasons for increasing interest in intrinsic techniques. We further summarize some practical issues of registration which include initializations and evaluations, and discuss some of our own observations, insights and foreseeable research trends.

Deformable Model Retrieval Based on Topological and Geometric Signatures

With the increasing popularity of 3D applications such as computer games, a lot of 3D geometry models are being created. To encourage sharing and reuse, techniques that support matching and retrieval of these models are emerging. However, only a few of them can handle deformable models, that is, models of different poses, and these methods are generally very slow. In this paper, we present a novel method for efficient matching and retrieval of 3D deformable models. Our research idea stresses using both topological and geometric features at the same time. First, we propose Topological Point Ring (TPR) analysis to locate reliable topological points and rings. Second, we capture both local and global geometric information to characterize each of these topological features. To compare the similarity of two models, we adapt the Earth Mover Distance (EMD) as the distance function and construct an indexing tree to accelerate the retrieval process. We demonstrate the performance of the new method, both in terms of accuracy and speed, through a large number of experiments.

Facial expression recognition in dynamic sequences

Automatic facial expression analysis aims to analyse human facial expressions and classify them into discrete categories. Methods based on existing work are reliant on extracting information from video sequences and employ either some form of subjective thresholding of dynamic information or attempt to identify the particular individual frames in which the expected behaviour occurs. These methods are inefficient as they require either additional subjective information, tedious manual work or fail to take advantage of the information contained in the dynamic signature from facial movements for the task of expression recognition.In this paper, a novel framework is proposed for automatic facial expression analysis which extracts salient information from video sequences but does not rely on any subjective preprocessing or additional user-supplied information to select frames with peak expressions. The experimental framework demonstrates that the proposed method outperforms static expression recognition systems in terms of recognition rate. The approach does not rely on action units (AUs), and therefore, eliminates errors which are otherwise propagated to the final result due to incorrect initial identification of AUs. The proposed framework explores a parametric space of over 300 dimensions and is tested with six state-of-the-art machine learning techniques. Such robust and extensive experimentation provides an important foundation for the assessment of the performance for future work. A further contribution of the paper is offered in the form of a user study. This was conducted in order to investigate the correlation between human cognitive systems and the proposed framework for the understanding of human emotion classification and the reliability of public databases. HighlightsExtraction of dynamic signals via a parametric space to improve the automatic facial expression recognition rate.An objective comparison with systems utilizing static apex expression recognition.The use of a visualisation technique for the analysis and initial understanding of facial feature data.An intuitive user study to investigate the correlation between human perception and machine vision.

Shape Retrieval of Non-rigid 3D Human Models

Abstract3D models of humans are commonly used within computer graphics and vision, and so the ability to distinguish between body shapes is an important shape retrieval problem. We extend our recent paper which provided a benchmark for testing non-rigid 3D shape retrieval algorithms on 3D human models. This benchmark provided a far stricter challenge than previous shape benchmarks. We have added 145 new models for use as a separate training set, in order to standardise the training data used and provide a fairer comparison. We have also included experiments with the FAUST dataset of human scans. All participants of the previous benchmark study have taken part in the new tests reported here, many providing updated results using the new data. In addition, further participants have also taken part, and we provide extra analysis of the retrieval results. A total of 25 different shape retrieval methods are compared.

Diffusion pruning for rapidly and robustly selecting global correspondences using local isometry

Finding correspondences between two surfaces is a fundamental operation in various applications in computer graphics and related fields. Candidate correspondences can be found by matching local signatures, but as they only consider local geometry, many are globally inconsistent. We provide a novel algorithm to prune a set of candidate correspondences to those most likely to be globally consistent. Our approach can handle articulated surfaces, and ones related by a deformation which is globally nonisometric, provided that the deformation is locally approximately isometric. Our approach uses an efficient diffusion framework, and only requires geodesic distance calculations in small neighbourhoods, unlike many existing techniques which require computation of global geodesic distances. We demonstrate that, for typical examples, our approach provides significant improvements in accuracy, yet also reduces time and memory costs by a factor of several hundred compared to existing pruning techniques. Our method is furthermore insensitive to holes, unlike many other methods.

An Analysis of Machine- and Human-Analytics in Classification

In this work, we present a study that traces the technical and cognitive processes in two visual analytics applications to a common theoretic model of soft knowledge that may be added into a visual analytics process for constructing a decision-tree model. Both case studies involved the development of classification models based on the “bag of features” approach. Both compared a visual analytics approach using parallel coordinates with a machine-learning approach using information theory. Both found that the visual analytics approach had some advantages over the machine learning approach, especially when sparse datasets were used as the ground truth. We examine various possible factors that may have contributed to such advantages, and collect empirical evidence for supporting the observation and reasoning of these factors. We propose an information-theoretic model as a common theoretic basis to explain the phenomena exhibited in these two case studies. Together we provide interconnected empirical and theoretical evidence to support the usefulness of visual analytics.

Visualization of time-series data in parameter space for understanding facial dynamics

Over the past decade, computer scientists and psychologists have made great efforts to collect and analyze facial dynamics data that exhibit different expressions and emotions. Such data is commonly captured as videos and are transformed into feature‐based time‐series prior to any analysis. However, the analytical tasks, such as expression classification, have been hindered by the lack of understanding of the complex data space and the associated algorithm space. Conventional graph‐based time‐series visualization is also found inadequate to support such tasks. In this work, we adopt a visual analytics approach by visualizing the correlation between the algorithm space and our goal – classifying facial dynamics. We transform multiple feature‐based time‐series for each expression in measurement space to a multi‐dimensional representation in parameter space. This enables us to utilize parallel coordinates visualization to gain an understanding of the algorithm space, providing a fast and cost‐effective means to support the design of analytical algorithms.

Embedding Retrieval of Articulated Geometry Models

Due to the popularity of computer games and animation, research on 3D articulated geometry model retrieval has attracted a lot of attention in recent years. However, most existing works extract high-dimensional features to represent models and suffer from practical limitations. First, misalignment in high-dimensional features may produce unreliable euclidean distances and affect retrieval accuracy. Second, the curse of dimensionality also degrades efficiency. In this paper, we propose an embedding retrieval framework to improve the practicability of these methods. It is based on a manifold learning technique, the Diffusion Map (DM). We project all pairwise distances onto a low-dimensional space. This improves retrieval accuracy because intercluster distances are exaggerated. Then we adapt the Density-Weighted Nyström extension and further propose a novel step to locally align the Nyström embedding to the eigensolver embedding so as to reduce extension error and preserve retrieval accuracy. Finally, we propose a heuristic to handle disconnected manifolds by augmenting the kernel matrix with multiple similarity measures and shortcut edges, and further discuss the choice of DM parameters. We have incorporated two existing matching algorithms for testing. Our experimental results show improvement in precision at high recalls and in speed. Our work provides a robust retrieval framework for the matching of multimedia data that lie on manifolds.

Motion Retrieval Based on Energy Morphing

Matching and retrieval of motion sequences has become an important research area in recent years, due to the increasing availability and popularity of motion capture data. The main challenge in matching two motion sequences is the diversity of the captured motions, including variable length, local shifting, local and global scaling. Most existing methods employ Dynamic Time Warping (DTW) or Uniform Scaling to handle these problems. In this paper, we propose a novel content-based method for matching of this human motion captured data. We convert the matching problem of motion capture data into a transportation problem. To solve this problem efficiently, we employ Earth Movers Distance (EMD) as the matching framework. To penalize any strayed matching, we provide a ground distance that works similar to Sakoe- Chiba band of DTW. Empirical results obtained are encouraging.

Automatic Aortic Root Segmentation with Shape Constraints and Mesh Regularisation

Fully automated 3D segmentation is not only challenging due to, for instance, ambiguities in appearance, but it is also computationally demanding. We present a fullyautomatic, learning-based deformable modelling method for segmenting the aortic root in CT images using a two-stage mesh deformation: a non-iterative boundary segmentation with a statistical shape model for shape constraint, followed by an iterative boundary refinement process. At both stages, we introduce a B-spline mesh regularisation technique to avoid mesh entanglement during deformation. The initialisation of the deformable model is achieved through efficient detection and localisation of the aortic root using marginal space learning, which carries out similarity parameter estimation in an incremental fashion. Quantitative comparisons are carried out against a state-of-the-art deformable model-based approach and an active shape model based segmentation. The proposed method achieves both a lower average mesh error of 1.39±0.29mm, and Hausdorff distance of 6.75±2.05mm. Compared to these two approaches, it results in much more regularised mesh surfaces with no tangled mesh faces.