Data that is represented with high dimensionality presents a computational complexity challenge for many existing algorithms. Limiting dimensionality by discarding attributes is sometimes a poor solution to this problem because significant high-level concepts may be encoded in the data across many or all of the attributes. Non-linear dimensionality reduction (NLDR) techniques have been successful with many problems at minimizing dimensionality while preserving intrinsic high-level concepts that are encoded with varying combinations of attributes. Unfortunately, many challenges remain with existing NLDR techniques, including excessive computational requirements, an inability to benefit from prior knowledge, and an inability to handle certain difficult conditions that occur in data with many real-world problems. Further, certain practical factors have limited advancement in NLDR, such as a lack of clarity regarding suitable applications for NLDR, and a general inavailability of efficient implementations of complex algorithms. This dissertation presents a collection of papers that advance the state of NLDR in each of these areas. Contributions of this dissertation include: • An NLDR algorithm, called Manifold Sculpting, that optimizes its solution using graduated optimization. This approach enables it to obtain better results than methods that only optimize an approximate problem. Additionally, Manifold Sculpting can benefit from prior knowledge about the problem. • An intelligent neighbor-finding technique called SAFFRON that improves the breadth of problems that existing NLDR techniques can handle. • A neighborhood refinement technique called CycleCut that further increases the robustness of existing NLDR techniques, and that can work in conjunction with SAFFRON to solve difficult problems. • Demonstrations of specific applications for NLDR techniques, including the estimation of state within dynamical systems, training of recurrent neural networks, and imputing missing values in data. • An open source toolkit containing each of the techniques described in this dissertation, as well as several existing NLDR algorithms, and other useful machine learning methods.

Single view reconstruction is fundamentally an under-constrained problem. We aim to develop new approaches to model human face and motion with model priors that restrict the space of possible solutions. First, we develop a novel approach to recover the 3D shape from a single view image under challenging conditions, such as large variations in illumination and pose. The problem is addressed by employing the techniques of non-linear manifold embedding and alignment. Specifically, the local image models for each patch of facial images and the local surface models for each patch of 3D shape are learned using a non-linear dimensionality reduction technique, and the correspondences between these local models are then learned by a manifold alignment method. Local models successfully remove the dependency of large training databases for human face modeling. By combining the local shapes, the global shape of a face can be reconstructed directly from a single linear system of equations via least square. Unfortunately, this learning-based approach cannot be successfully applied to the problem of human motion modeling due to the internal and external variations in single view video-based marker-less motion capture. Therefore, we introduce a new model-based approach for capturing human motion using a stream of depth images from a single depth sensor. While a depth sensor provides metric 3D information, using a single sensor, instead of a camera array, results in a view-dependent and incomplete measurement of object motion. We develop a novel two-stage template fitting algorithm that is invariant to subject size and view-point variations, and robust to occlusions. Starting from a known pose, our algorithm first estimates a body configuration through temporal registration, which is used to search the template motion database for a best match. The best match body configuration as well as its corresponding surface mesh model are deformed to fit the input depth map, filling in the part that is occluded from the input and compensating for differences in pose and body-size between the input image and the template. Our approach does not require any makers, user-interaction, or appearance-based tracking. Experiments show that our approaches can achieve good modeling results for human face and motion, and are capable of dealing with variety of challenges in single view reconstruction, e.g., occlusion.

Τμηματοποίηση υφής ονομάζεται η διαδικασία του διαμερισμού μίας εικόνας σε πολλαπλά τμήματα-περιοχές, με κριτήριο την υφή κάθε περιοχής. Η διαδικασία αυτή βρίσκει πολλές εφαρμογές στους τομείς της υπολογιστικής όρασης, της ανάκτησης εικόνων, της ρομποτικής, της ανάλυσης δορυφορικών εικόνων κλπ. Αντικείμενο της παρούσης εργασίας είναι να διερευνηθεί η ικανότητα των αλγορίθμων μη γραμμικής ελάττωσης διάστασης, και ιδιαίτερα του αλγορίθμου Laplacian Eigenmaps, να παράγει μία αποδοτική αναπαράσταση των δεδομένων που προέρχονται από πολυφασματική ανάλυση εικόνων με χρήση φίλτρων Gabor, για την επίλυση του προβλήματος της τμηματοποίησης εικόνων υφής. Για το σκοπό αυτό προτείνεται μία νέα μέθοδος επιβλεπόμενης τμηματοποίησης υφής, που αξιοποιεί μία χαμηλής διάστασης αναπαράσταση των χαρακτηριστικών διανυσμάτων, και γνωστούς αλγόριθμους ομαδοποίησης δεδομένων όπως οι Fuzzy C-means και K-means, για την παραγωγή της τελικής τμηματοποίησης. Η αποτελεσματικότητα της μεθόδου συγκρίνεται με παρόμοιες μεθόδους που έχουν προταθεί στη βιβλιογραφία, και χρησιμοποιούν την αρχική , υψηλών διαστάσεων, αναπαράσταση των χαρακτηριστικών διανυσμάτων. Τα πειράματα διενεργήθηκαν χρησιμοποιώντας την βάση εικόνων υφής Brodatz. Κατά το στάδιο αξιολόγησης της μεθόδου, χρησιμοποιήθηκε ο δείκτης Rand index σαν μέτρο ομοιότητας ανάμεσα σε κάθε παραγόμενη τμηματοποίηση και την αντίστοιχη ground-truth τμηματοποίηση. / Texture segmentation is the process of partitioning an image into multiple segments (regions) based on their texture, with many applications in the area of computer vision, image retrieval, robotics, satellite imagery etc. The objective of this thesis is to investigate the ability of non-linear dimensionality reduction algorithms, and especially of LE algorithm, to produce an efficient representation for data derived from multi-spectral image analysis using Gabor filters, in solving the texture segmentation problem. For this purpose, we introduce a new supervised texture segmentation algorithm, which exploits a low-dimensional representation of feature vectors and well known clustering methods, such as Fuzzy C-means and K-means, to produce the final segmentation. The effectiveness of this method was compared to that of similar methods proposed in the literature, which use the initial high-dimensional representation of feature vectors. Experiments were performed on Brodatz texture database. During evaluation stage, Rand index has been used as a similarity measure between each segmentation and the corresponding ground-truth segmentation.