Current Search: Foroosh, Hassan (x)
View All Items
(41 - 45 of 45)
Pages
- Title
- Mathematical and Computational Methods for Freeform Optical Shape Description.
- Creator
-
Kaya, Ilhan, Foroosh, Hassan, Rolland, Jannick, Turgut, Damla, Thompson, Kevin, Ilegbusi, Olusegun, University of Central Florida
- Abstract / Description
-
Slow-servo single-point diamond turning as well as advances in computer controlled small lap polishing enable the fabrication of freeform optics, specifically, optical surfaces for imaging applications that are not rotationally symmetric. Freeform optical elements will have a profound importance in the future of optical technology. Orthogonal polynomials added onto conic sections have been extensively used to describe optical surface shapes. The optical testing industry has chosen to...
Show moreSlow-servo single-point diamond turning as well as advances in computer controlled small lap polishing enable the fabrication of freeform optics, specifically, optical surfaces for imaging applications that are not rotationally symmetric. Freeform optical elements will have a profound importance in the future of optical technology. Orthogonal polynomials added onto conic sections have been extensively used to describe optical surface shapes. The optical testing industry has chosen to represent the departure of a wavefront under test from a reference sphere in terms of orthogonal ?-polynomials, specifically Zernike polynomials. Various forms of polynomials for describing freeform optical surfaces may be considered, however, both in optical design and in support of fabrication. More recently, radial basis functions were also investigated for optical shape description. In the application of orthogonal ?-polynomials to optical freeform shape description, there are important limitations, such as the number of terms required as well as edge-ringing and ill-conditioning in representing the surface with the accuracy demanded by most stringent optics applications. The first part of this dissertation focuses upon describing freeform optical surfaces with ? polynomials and shows their limitations when including higher orders together with possible remedies. We show that a possible remedy is to use edge clustered-fitting grids. Provided different grid types, we furthermore compared the efficacy of using different types of ? polynomials, namely Zernike and gradient orthogonal Q polynomials. In the second part of this thesis, a local, efficient and accurate hybrid method is developed in order to greatly reduce the order of polynomial terms required to achieve higher level of accuracy in freeform shape description that were shown to require thousands of terms including many higher order terms under prior art. This comes at the expense of multiple sub-apertures, and as such computational methods may leverage parallel processing. This new method combines the assets of both radial basis functions and orthogonal phi-polynomials for freeform shape description and is uniquely applicable across any aperture shape due to its locality and stitching principles. Finally in this thesis, in order to comprehend the possible advantages of parallel computing for optical surface descriptions, the benefits of making an effective use of impressive computational power offered by multi-core platforms for the computation of ?-polynomials are investigated. The ?-polynomials, specifically Zernike and gradient orthogonal Q-polynomials, are implemented with a set of recurrence based parallel algorithms on Graphics Processing Units (GPUs). The results show that more than an order of magnitude speedup is possible in the computation of ?-polynomials over a sequential implementation if the recurrence based parallel algorithms are adopted.
Show less - Date Issued
- 2013
- Identifier
- CFE0005012, ucf:49993
- Format
- Document (PDF)
- PURL
- http://purl.flvc.org/ucf/fd/CFE0005012
- Title
- Exploring sparsity, self-similarity, and low rank approximation in action recognition, motion retrieval, and action spotting.
- Creator
-
Sun, Chuan, Foroosh, Hassan, Hughes, Charles, Tappen, Marshall, Sukthankar, Rahul, Moshell, Jack, University of Central Florida
- Abstract / Description
-
This thesis consists of $4$ major parts. In the first part (Chapters $1-2$), we introduce the overview, motivation, and contribution of our works, and extensively survey the current literature for $6$ related topics. In the second part (Chapters $3-7$), we explore the concept of ``Self-Similarity" in two challenging scenarios, namely, the Action Recognition and the Motion Retrieval. We build three-dimensional volume representations for both scenarios, and devise effective techniques that can...
Show moreThis thesis consists of $4$ major parts. In the first part (Chapters $1-2$), we introduce the overview, motivation, and contribution of our works, and extensively survey the current literature for $6$ related topics. In the second part (Chapters $3-7$), we explore the concept of ``Self-Similarity" in two challenging scenarios, namely, the Action Recognition and the Motion Retrieval. We build three-dimensional volume representations for both scenarios, and devise effective techniques that can produce compact representations encoding the internal dynamics of data. In the third part (Chapter $8$), we explore the challenging action spotting problem, and propose a feature-independent unsupervised framework that is effective in spotting action under various real situations, even under heavily perturbed conditions. The final part (Chapters $9$) is dedicated to conclusions and future works.For action recognition, we introduce a generic method that does not depend on one particular type of input feature vector. We make three main contributions: (i) We introduce the concept of Joint Self-Similarity Volume (Joint SSV) for modeling dynamical systems, and show that by using a new optimized rank-1 tensor approximation of Joint SSV one can obtain compact low-dimensional descriptors that very accurately preserve the dynamics of the original system, e.g. an action video sequence; (ii) The descriptor vectors derived from the optimized rank-1 approximation make it possible to recognize actions without explicitly aligning the action sequences of varying speed of execution or difference frame rates; (iii) The method is generic and can be applied using different low-level features such as silhouettes, histogram of oriented gradients (HOG), etc. Hence, it does not necessarily require explicit tracking of features in the space-time volume. Our experimental results on five public datasets demonstrate that our method produces very good results and outperforms many baseline methods.For action recognition for incomplete videos, we determine whether incomplete videos that are often discarded carry useful information for action recognition, and if so, how one can represent such mixed collection of video data (complete versus incomplete, and labeled versus unlabeled) in a unified manner. We propose a novel framework to handle incomplete videos in action classification, and make three main contributions: (i) We cast the action classification problem for a mixture of complete and incomplete data as a semi-supervised learning problem of labeled and unlabeled data. (ii) We introduce a two-step approach to convert the input mixed data into a uniform compact representation. (iii) Exhaustively scrutinizing $280$ configurations, we experimentally show on our two created benchmarks that, even when the videos are extremely sparse and incomplete, it is still possible to recover useful information from them, and classify unknown actions by a graph based semi-supervised learning framework.For motion retrieval, we present a framework that allows for a flexible and an efficient retrieval of motion capture data in huge databases. The method first converts an action sequence into a self-similarity matrix (SSM), which is based on the notion of self-similarity. This conversion of the motion sequences into compact and low-rank subspace representations greatly reduces the spatiotemporal dimensionality of the sequences. The SSMs are then used to construct order-3 tensors, and we propose a low-rank decomposition scheme that allows for converting the motion sequence volumes into compact lower dimensional representations, without losing the nonlinear dynamics of the motion manifold. Thus, unlike existing linear dimensionality reduction methods that distort the motion manifold and lose very critical and discriminative components, the proposed method performs well, even when inter-class differences are small or intra-class differences are large. In addition, the method allows for an efficient retrieval and does not require the time-alignment of the motion sequences. We evaluate the performance of our retrieval framework on the CMU mocap dataset under two experimental settings, both demonstrating very good retrieval rates.For action spotting, our framework does not depend on any specific feature (e.g. HOG/HOF, STIP, silhouette, bag-of-words, etc.), and requires no human localization, segmentation, or framewise tracking. This is achieved by treating the problem holistically as that of extracting the internal dynamics of video cuboids by modeling them in their natural form as multilinear tensors. To extract their internal dynamics, we devised a novel Two-Phase Decomposition (TP-Decomp) of a tensor that generates very compact and discriminative representations that are robust to even heavily perturbed data. Technically, a Rank-based Tensor Core Pyramid (Rank-TCP) descriptor is generated by combining multiple tensor cores under multiple ranks, allowing to represent video cuboids in a hierarchical tensor pyramid. The problem then reduces to a template matching problem, which is solved efficiently by using two boosting strategies: (i) to reduce the search space, we filter the dense trajectory cloud extracted from the target video; (ii) to boost the matching speed, we perform matching in an iterative coarse-to-fine manner. Experiments on 5 benchmarks show that our method outperforms current state-of-the-art under various challenging conditions. We also created a challenging dataset called Heavily Perturbed Video Arrays (HPVA) to validate the robustness of our framework under heavily perturbed situations.
Show less - Date Issued
- 2014
- Identifier
- CFE0005554, ucf:50290
- Format
- Document (PDF)
- PURL
- http://purl.flvc.org/ucf/fd/CFE0005554
- Title
- Applications of Computer Vision Technologies of Automated Crack Detection and Quantification for the Inspection of Civil Infrastructure Systems.
- Creator
-
Wu, Liuliu, Yun, Hae-Bum, Nam, Boo Hyun, Catbas, Necati, Foroosh, Hassan, University of Central Florida
- Abstract / Description
-
Many components of existing civil infrastructure systems, such as road pavement, bridges, and buildings, are suffered from rapid aging, which require enormous nation's resources from federal and state agencies to inspect and maintain them. Crack is one of important material and structural defects, which must be inspected not only for good maintenance of civil infrastructure with a high quality of safety and serviceability, but also for the opportunity to provide early warning against failure....
Show moreMany components of existing civil infrastructure systems, such as road pavement, bridges, and buildings, are suffered from rapid aging, which require enormous nation's resources from federal and state agencies to inspect and maintain them. Crack is one of important material and structural defects, which must be inspected not only for good maintenance of civil infrastructure with a high quality of safety and serviceability, but also for the opportunity to provide early warning against failure. Conventional human visual inspection is still considered as the primary inspection method. However, it is well established that human visual inspection is subjective and often inaccurate. In order to improve current manual visual inspection for crack detection and evaluation of civil infrastructure, this study explores the application of computer vision techniques as a non-destructive evaluation and testing (NDE(&)T) method for automated crack detection and quantification for different civil infrastructures. In this study, computer vision-based algorithms were developed and evaluated to deal with different situations of field inspection that inspectors could face with in crack detection and quantification. The depth, the distance between camera and object, is a necessary extrinsic parameter that has to be measured to quantify crack size since other parameters, such as focal length, resolution, and camera sensor size are intrinsic, which are usually known by camera manufacturers. Thus, computer vision techniques were evaluated with different crack inspection applications with constant and variable depths. For the fixed-depth applications, computer vision techniques were applied to two field studies, including 1) automated crack detection and quantification for road pavement using the Laser Road Imaging System (LRIS), and 2) automated crack detection on bridge cables surfaces, using a cable inspection robot. For the various-depth applications, two field studies were conducted, including 3) automated crack recognition and width measurement of concrete bridges' cracks using a high-magnification telescopic lens, and 4) automated crack quantification and depth estimation using wearable glasses with stereovision cameras.From the realistic field applications of computer vision techniques, a novel self-adaptive image-processing algorithm was developed using a series of morphological transformations to connect fragmented crack pixels in digital images. The crack-defragmentation algorithm was evaluated with road pavement images. The results showed that the accuracy of automated crack detection, associated with artificial neural network classifier, was significantly improved by reducing both false positive and false negative. Using up to six crack features, including area, length, orientation, texture, intensity, and wheel-path location, crack detection accuracy was evaluated to find the optimal sets of crack features. Lab and field test results of different inspection applications show that proposed compute vision-based crack detection and quantification algorithms can detect and quantify cracks from different structures' surface and depth. Some guidelines of applying computer vision techniques are also suggested for each crack inspection application.
Show less - Date Issued
- 2015
- Identifier
- CFE0005743, ucf:50089
- Format
- Document (PDF)
- PURL
- http://purl.flvc.org/ucf/fd/CFE0005743
- Title
- Complex Affect Recognition in the Wild.
- Creator
-
Nojavanasghari, Behnaz, Hughes, Charles, Morency, Louis-Philippe, Sukthankar, Gita, Foroosh, Hassan, Morency, Louis-Philippe, University of Central Florida
- Abstract / Description
-
Arti?cial social intelligence is a step towards human-like human-computer interaction. One important milestone towards building socially intelligent systems is enabling computers with the ability to process and interpret the social signals of humans in the real world. Social signals include a wide range of emotional responses from a simple smile to expressions of complex affects.This dissertation revolves around computational models for social signal processing in the wild, using multimodal...
Show moreArti?cial social intelligence is a step towards human-like human-computer interaction. One important milestone towards building socially intelligent systems is enabling computers with the ability to process and interpret the social signals of humans in the real world. Social signals include a wide range of emotional responses from a simple smile to expressions of complex affects.This dissertation revolves around computational models for social signal processing in the wild, using multimodal signals with an emphasis on the visual modality. We primarily focus on complex affect recognition with a strong interest in curiosity. In this dissertation,we ?rst present our collected dataset, EmoReact. We provide detailed multimodal behavior analysis across audio-visual signals and present unimodal and multimodal classi?cation models for affect recognition. Second, we present a deep multimodal fusion algorithm to fuse information from visual, acoustic and verbal channels to achieve a uni?ed classi?cation result. Third, we present a novel system to synthesize, recognize and localize facial occlusions. The proposed framework is based on a three-stage process: 1) Synthesis of naturalistic occluded faces, which include hand over face occlusions as well as other common occlusions such as hair bangs, scarf, hat, etc. 2) Recognition of occluded faces and differentiating between hand over face and other types of facial occlusions. 3) Localization of facial occlusions and identifying the occluded facial regions. Region of facial occlusion, plays an importantroleinrecognizingaffectandashiftinlocationcanresultinaverydifferentinterpretation, e.g., hand over chin can indicate contemplation, while hand over eyes may show frustration or sadness. Finally, we show the importance of considering facial occlusion type and region in affect recognition through achieving promising results in our experiments.
Show less - Date Issued
- 2017
- Identifier
- CFE0007291, ucf:52163
- Format
- Document (PDF)
- PURL
- http://purl.flvc.org/ucf/fd/CFE0007291
- Title
- Signal processing with Fourier analysis, novel algorithms and applications.
- Creator
-
Syed, Alam, Foroosh, Hassan, Sun, Qiyu, Bagci, Ulas, Rahnavard, Nazanin, Atia, George, Katsevich, Alexander, University of Central Florida
- Abstract / Description
-
Fourier analysis is the study of the way general functions may be represented or approximatedby sums of simpler trigonometric functions, also analogously known as sinusoidal modeling. Theoriginal idea of Fourier had a profound impact on mathematical analysis, physics, and engineeringbecause it diagonalizes time-invariant convolution operators. In the past signal processing was atopic that stayed almost exclusively in electrical engineering, where only the experts could cancelnoise, compress...
Show moreFourier analysis is the study of the way general functions may be represented or approximatedby sums of simpler trigonometric functions, also analogously known as sinusoidal modeling. Theoriginal idea of Fourier had a profound impact on mathematical analysis, physics, and engineeringbecause it diagonalizes time-invariant convolution operators. In the past signal processing was atopic that stayed almost exclusively in electrical engineering, where only the experts could cancelnoise, compress and reconstruct signals. Nowadays it is almost ubiquitous, as everyone now dealswith modern digital signals.Medical imaging, wireless communications and power systems of the future will experience moredata processing conditions and wider range of applications requirements than the systems of today.Such systems will require more powerful, efficient and flexible signal processing algorithms thatare well designed to handle such needs. No matter how advanced our hardware technology becomeswe will still need intelligent and efficient algorithms to address the growing demands in signalprocessing. In this thesis, we investigate novel techniques to solve a suite of four fundamentalproblems in signal processing that have a wide range of applications. The relevant equations, literatureof signal processing applications, analysis and final numerical algorithms/methods to solvethem using Fourier analysis are discussed for different applications in the electrical engineering /computer science. The first four chapters cover the following topics of central importance in thefield of signal processing: Fast Phasor Estimation using Adaptive Signal Processing (Chapter 2) Frequency Estimation from Nonuniform Samples (Chapter 3) 2D Polar and 3D Spherical Polar Nonuniform Discrete Fourier Transform (Chapter 4)iv Robust 3D registration using Spherical Polar Discrete Fourier Transform and Spherical Harmonics(Chapter 5)Even though each of these four methods discussed may seem completely disparate, the underlyingmotivation for more efficient processing by exploiting the Fourier domain signal structureremains the same. The main contribution of this thesis is the innovation in the analysis, synthesis, discretization of certain well-known problems like phasor estimation, frequency estimation, computations of a particular non-uniform Fourier transform and signal registration on the transformed domain. We conduct propositions and evaluations of certain applications relevant algorithms suchas, frequency estimation algorithm using non-uniform sampling, polar and spherical polar Fourier transform. The techniques proposed are also useful in the field of computer vision and medical imaging. From a practical perspective, the proposed algorithms are shown to improve the existing solutions in the respective fields where they are applied/evaluated. The formulation and final proposition is shown to have a variety of benefits. Future work with potentials in medical imaging, directional wavelets, volume rendering, video/3D object classifications, high dimensional registration are also discussed in the final chapter. Finally, in the spirit of reproducible research, we release the implementation of these algorithms to the public using Github.
Show less - Date Issued
- 2017
- Identifier
- CFE0006803, ucf:51775
- Format
- Document (PDF)
- PURL
- http://purl.flvc.org/ucf/fd/CFE0006803
