Research Activities at Florida State Vision Group Florida State University Xiuwen Liu Department of Computer Science Florida State University http://fsvision.fsu.edu http://www.cs.fsu.edu/~liux/courses/intro-seminar-08.ppt Research Statement My research goal is to create machines that can “see” with similar and super human performance and their applications • This seems a trivial problem as each of us can do this without any effort • Computer + Camera = “A See Machine” ? 10/13/2008 10:16:20 PM intro-seminar-08.ppt Visual Pathway 10/13/2008 10:16:21 PM intro-seminar-08.ppt Visual Illusion 10/13/2008 10:16:21 PM intro-seminar-08.ppt Outline Motivations • Some applications of computer vision and pattern recognition techniques Some of my research projects Related courses Contact information 10/13/2008 10:16:22 PM intro-seminar-08.ppt Computer Vision Applications 10/13/2008 10:16:23 PM intro-seminar-08.ppt DARPA Urban Challenge U.S. Congressional set a goal that by 2015 one-third of the U.S. Armed Forces’ operational ground combat vehicles be unmanned • Section 220 of the Floyd D. Spence National Defense Authorization Act for Fiscal Year 2001, Public Law 106-398 The Urban Challenge • http://www.darpa.mil/grandchallenge/index.asp • Teams competed to build an autonomous vehicle able to complete a 60-mile urban course safely in less than 6 hours 10/13/2008 10:16:23 PM intro-seminar-08.ppt 7 DARPA Urban Challenge 10/13/2008 10:16:24 PM intro-seminar-08.ppt 8 3D Urban Models 10/13/2008 10:16:25 PM intro-seminar-08.ppt 9 10/13/2008 10:16:26 PM intro-seminar-08.ppt 10 10/13/2008 10:16:26 PM intro-seminar-08.ppt 11 Computational Photography 10/13/2008 10:16:27 PM intro-seminar-08.ppt 12 Face Detection for Auto Focusing Some cameras now have the built-in algorithms to automatically detect and focus on the faces as they are the most important subject in everyday photographing Source: http://facedetection.fujifilmusa.com/ 10/13/2008 10:16:28 PM intro-seminar-08.ppt 13 Human-Computer Interactions 10/13/2008 10:16:28 PM intro-seminar-08.ppt Sign Language Recognition 10/13/2008 10:16:32 PM intro-seminar-08.ppt CyberKnife 10/13/2008 10:16:34 PM intro-seminar-08.ppt CyberKnife – Cont. 10/13/2008 10:16:35 PM intro-seminar-08.ppt Image-Guided Neurosurgery 10/13/2008 10:16:36 PM intro-seminar-08.ppt Intelligent Transportation Systems http://dfwtraffic.dot.state.tx.us/dal-cam-nf.asp 10/13/2008 10:16:37 PM intro-seminar-08.ppt Computer Vision Applications – cont. Military applications • Automated target recognition 10/13/2008 10:16:37 PM intro-seminar-08.ppt Biometrics Iris code can achieve zero false acceptance intro-seminar-08.ppt 10/13/2008 10:16:38 PM Computer Vision in Sports How was the yellow created? 10/13/2008 10:16:39 PM intro-seminar-08.ppt Generic Image Modeling How can we characterize all these images perceptually? 10/13/2008 10:16:39 PM intro-seminar-08.ppt Spectral Histogram Representation Spectral histogram • Given a bank of filters F(a), a = 1, …, K, a spectral histogram is defined as the marginal distribution of filter responses I(a ) (v) F (a ) * I(v) H (a ) I 1 (a ) ( z) δ ( z I (v)) |I| v H I ( H I(1) , H I( 2) ,, H I( K ) ) 10/13/2008 10:16:40 PM intro-seminar-08.ppt Spectral Histogram Representation - continued Choice • • • • of filters Laplacian of Gaussian filters Gabor filters Gradient filters Intensity filter LoG filter 10/13/2008 10:16:40 PM Gabor filter intro-seminar-08.ppt Spectral Histogram Representation - continued 10/13/2008 10:16:41 PM intro-seminar-08.ppt Texture Synthesis Examples - continued Observed image An Synthesized image image with periodic structures 10/13/2008 10:16:41 PM intro-seminar-08.ppt Object Synthesis Examples - continued 10/13/2008 10:16:42 PM intro-seminar-08.ppt Performance Comparison 10/13/2008 10:16:42 PM intro-seminar-08.ppt Face Detection Based On Spectral Representations Face detection is to detect all instances of faces in a given image Each image window is represented by its spectral histogram • A support vector machine is trained on training faces • Then the trained support vector machine is used to classify each image window in an input image More results at http://fsvision.fsu.edu/face-detection 10/13/2008 10:16:43 PM intro-seminar-08.ppt Face detection - continued 10/13/2008 10:16:43 PM intro-seminar-08.ppt Face detection - continued 10/13/2008 10:16:44 PM intro-seminar-08.ppt Face detection - continued 10/13/2008 10:16:45 PM intro-seminar-08.ppt Rotation Invariant Face Detection 10/13/2008 10:16:45 PM intro-seminar-08.ppt Rotation Invariant Face Detection - continued 10/13/2008 10:16:46 PM intro-seminar-08.ppt Linear Representations Linear representations are widely used in appearance-based object recognition and other applications • Simple to implement and analyze • Efficient to compute • Effective for many applications a ( I ,U ) U I R T 10/13/2008 10:16:46 PM d intro-seminar-08.ppt Standard Linear Representations Principal Component Analysis • Designed to minimize the reconstruction error on the training set • Obtained by calculating eigenvectors of the co-variance matrix Fisher Discriminant Analysis • Designed to maximize the separation between means of each class • Obtained by solving a generalized eigen problem Independent Component Analysis • Designed to maximize the statistical independence among coefficients along different directions • Obtained by solving an optimization problem with some object function such as mutual information, negentropy, .... 10/13/2008 10:16:47 PM intro-seminar-08.ppt Standard Linear Representations - continued Standard linear representations are sub optimal for recognition applications • Evidence in the literature • A toy example – Standard representations give the worst recognition performance Optimal 10/13/2008 10:16:48 PM component analysis intro-seminar-08.ppt Performance Measure - continued Suppose there are C classes to be recognized • Each class has ktrain training images • It has kcross cross validation images • We used h(x) = 1/(1+exp(-2bx) 10/13/2008 10:16:48 PM intro-seminar-08.ppt Performance Measure - continued F(U) depends on the span of U but is invariant to change of basis • In other words, F(U)=F(UO) for any orthonormal matrix O • The search space of F(U) is the set of all the subspaces, which is known as the Grassmann manifold – It is not a flat vector space and gradient flow must take the underlying geometry of the manifold into account 10/13/2008 10:16:49 PM intro-seminar-08.ppt Deterministic Gradient Flow - continued Gradient at [J] (first d columns of n x n identity matrix) 10/13/2008 10:16:49 PM intro-seminar-08.ppt Deterministic Gradient Flow - continued Gradient at U: Compute Q such that QU=J Deterministic gradient flow on Grassmann manifold 10/13/2008 10:16:50 PM intro-seminar-08.ppt Stochastic Gradient and Updating Rules Stochastic gradient is obtained by adding a stochastic component Discrete updating rules 10/13/2008 10:16:50 PM intro-seminar-08.ppt MCMC Simulated Annealing Optimization Algorithm Let X(0) be any initial condition and t=0 1. 2. 3. 4. 5. 6. 7. Calculate the gradient matrix A(Xt) Generate d(n-d) independent realizations of wij’s Compute Y (Xt+1) according to the updating rules Compute F(Y) and F(Xt) and set dF=F(Y)- F(Xt) Set Xt+1 = Y with probability min{exp(dF/Dt),1} Set Dt+1 = Dt / g and set t=t+1 Go to step 1 10/13/2008 10:16:50 PM intro-seminar-08.ppt ORL Face Dataset 10/13/2008 10:16:51 PM intro-seminar-08.ppt Performance Comparison 10/13/2008 10:16:53 PM intro-seminar-08.ppt Performance Comparison – cont. 10/13/2008 10:16:55 PM intro-seminar-08.ppt Brain Curve Classification 10/13/2008 10:16:56 PM intro-seminar-08.ppt Brain Curve Classification – cont. 10/13/2008 10:16:57 PM intro-seminar-08.ppt Real-time Scene Interpretation Object detection and recognition problem • Given a set of images, find regions in these images which contain instances of relevant objects • Here the number of relevant objects is assumed to be large – For example, the system should be able to handle 30,000 different kinds of objects, an estimate of the human brain’s capacity for basic level visual categorization [I. Biederman, Psychological Review, vol. 94, pp. 115-147, 1987] 10/13/2008 10:16:58 PM intro-seminar-08.ppt Global Monitoring Through High-resolution Satellite Images 10/13/2008 10:16:58 PM intro-seminar-08.ppt Problem Statement for Scene Interpretation Object detection and recognition problem • Given a set of images, find regions in these images which contain instances of relevant objects • Here the number of relevant objects is assumed to be large – For example, the system should be able to handle 30,000 different kinds of objects, an estimate of the human’s capacity for basic level visual categorization [I. Biederman, Psychological Review, vol. 94, pp. 115-147, 1987] Goal • Develop a system that can achieve real-time detection and recognition for images of size 640 x 480 with high accuracy – Say, at a frame rate of 15 frames per second 10/13/2008 10:16:59 PM intro-seminar-08.ppt Existing Approaches Fast methods but low accuracy • One can for example classify one pixel at a time • However, it is to identify airplanes with high accuracy due to high false positives and negatives 10/13/2008 10:17:00 PM intro-seminar-08.ppt Existing Approaches – cont. Fast methods but low accuracy • One can for example classify one pixel at a time • However, it is to identify airplanes with high accuracy Methods with good accuracy but slow • One can in theory use deformable template matching to locate instances of airplanes • It may need several hours to process one image 10/13/2008 10:17:00 PM intro-seminar-08.ppt Proposed Framework 10/13/2008 10:17:01 PM intro-seminar-08.ppt Specifications and Requirements We want to detect and recognize at least 30,000 object classes in images • At four different scales • Using exhaustive search of local windows, that is, we do not assume segmentation or other pre-processing • If we assume objects are in some (e.g. 21 x 21) windows, this means that there will be many (18,432,000) local windows to be classified/processed • We want to do this on a 3.6 Ghz Dell Precision workstation with an estimated performance of 28,665.4 MIPS • This amounts to that we have about 1555 instructions to process a 21 x 21 local window 10/13/2008 10:17:01 PM intro-seminar-08.ppt Requirements – cont. To achieve the specifications, we need two critical components • A classifier that can reduce the average classification time effectively – Note that on average we have 1555 instructions; if we can process 90% of those windows using only 100 instructions per window, we can have on average 14,650 instructions for the remaining 10% local windows • Features that can discriminate a large number of objects and can be computed using a few instructions – Do such features exist? 10/13/2008 10:17:01 PM intro-seminar-08.ppt Topological Local Spectral Histograms We introduce a new class of features, which we called TLSH features • It is defined relative to a chosen set of filters • For a given filter, it is defined as a histogram of a local window of the filtered image • One bin of the histogram is given by 10/13/2008 10:17:02 PM intro-seminar-08.ppt Topological Local Spectral Histogram Example Convolution is implemented using FPGAs 10/12/2008 9:51:38 PM intro-seminar-08.ppt Local Spectral Histogram Features 10/12/2008 9:51:39 PM intro-seminar-08.ppt Field Programmable Gate Arrays • Two primary methods for computation • Hard Wired Application Specific Integrated Circuit (ASIC) • Software-programmed microprocessors • New Approach • Programmable hardware • Field Programmable Gate Arrays (FPGAs) represent a breakthrough in computing technology – Especially for intrinsically parallel applications 10/12/2008 9:51:41 PM intro-seminar-08.ppt μP/ ASIC / FPGA Comparison Summary μP ASIC FPGA Programmable (flexible) Fixed Design Functionality (inflexible) Programmable (flexible) Relatively Slow Serial Computation Very Fast, highly parallelized computation Fast, Parallel Computation Floating and Fixed Point Fixed Point / Floating Fixed Point / Floating Relatively Inexpensive Design Cycle (Software) Expensive Design Cycle (requires chip design) Relatively Inexpensive Design Cycle Limited Bandwidth Very High Bandwidth Near ASIC Bandwidth Standard High Level Languages C/C++ or Assembly Hardware Description Language for Design / Simulation VHDL / Verilog Hardware Description Language for Design / Simulation VHDL / Verilog 10/12/2008 9:51:44 PM intro-seminar-08.ppt Hardware vs. Software L 1 • Software Implementation: y (n ) xk (n ) hk k 0 Sum = 0.0 I = 0; While (I < L) tmp = x(i) * h(i) Sum = Sum + tmp I = I+1 end 10/12/2008 9:51:45 PM A typical software implementation takes 4*L instructions to compute one convolution intro-seminar-08.ppt Hardware vs. Software A custom hardware implementation Multiply/Accumulate performed in parallel Can be done in one clock cycle 10/12/2008 9:51:46 PM intro-seminar-08.ppt Convolution Timing Diagram Convolution Start Signal 10/12/2008 9:51:46 PM All nine response values finished Clock intro-seminar-08.ppt Every 7 Clock Cycles: 9 new response values Topological Local Spectral Histograms – cont. Why TLSH features? • It provides a very rich set of over-complete features – For example, suppose we have 22 filters, there will be 1,173,942 different TLSH features within a 21 x 21 region, considering different windows and different filters – TLSH features are more effective than Haar features used by Viola and Jones [P. Viola and M. Jones, International Journal of Computer Vision, vol. 57, pp. 137-154, 2004] 10/12/2008 9:51:47 PM intro-seminar-08.ppt ORL Face Dataset 10/12/2008 9:51:48 PM intro-seminar-08.ppt Comparison Between Haar and TLSH Features 10/12/2008 9:51:49 PM intro-seminar-08.ppt COIL Dataset 10/12/2008 9:51:49 PM intro-seminar-08.ppt Comparison Between Haar and TLSH Features 10/12/2008 9:51:50 PM intro-seminar-08.ppt Texture Dataset 10/12/2008 9:51:51 PM intro-seminar-08.ppt Comparison Between Haar and TLSH Features 10/12/2008 9:51:52 PM intro-seminar-08.ppt Mixed Dataset 10/12/2008 9:51:52 PM intro-seminar-08.ppt Comparison Between Haar and TLSH Features 10/12/2008 9:51:53 PM intro-seminar-08.ppt Comparison Between Haar and TLSH Features 10/12/2008 9:51:54 PM intro-seminar-08.ppt Classifier To achieve the specification, we also need a classifier that takes only a few instructions to make a decision on average • At the same time, we need to achieve high accuracy We propose to use a look-up table tree classifier • I.e., a decision tree classifier where each node is implemented by a look-up table 10/12/2008 9:51:54 PM intro-seminar-08.ppt Look-up Table Tree Classifier 10/12/2008 9:51:56 PM intro-seminar-08.ppt Look-up Table Tree Classifier 10/12/2008 9:51:57 PM intro-seminar-08.ppt An Example Path in a Decision Tree 10/12/2008 9:51:57 PM intro-seminar-08.ppt Constructing Look-up Table Decision Tree Joint optimization of clustering, TLSH features, and optimal linear projections • We want to maximize the separations between marginal distributions of different clusters • We can do the optimization iteratively – We can do clustering first using current TLSH features and projections to maximize the separations – We can find optimal TLSH features given linear projections – Then we can find optimal linear projections given updated TLSH features 10/12/2008 9:51:58 PM intro-seminar-08.ppt Performance Comparison RCT – Rapid Classification Tree, implemented by Keith Haynes 10/12/2008 9:51:58 PM intro-seminar-08.ppt Detection and Recognition 10/12/2008 9:51:59 PM intro-seminar-08.ppt Detection and Recognition 10/12/2008 9:52:00 PM intro-seminar-08.ppt Content-based Video Representation, Indexing and Retrieval A video is an extrinsic 3D representation of a 4D volume • 3D spatial space + 1D temporal space = 4D volume • For video, 2D image space + 1D temporal space = 3D volume Our group is working an intrinsic 4D representation for video • By first reconstructing the scene using SLAM (Simultaneous localization and mapping) and stereopsis 10/12/2008 9:52:18 PM intro-seminar-08.ppt Computer Vision on Mobile Devices Source: http://www.appliedmediaanalysis.com/MATES.htm 10/12/2008 9:52:20 PM intro-seminar-08.ppt Computer Vision for Gerotechnology As mobile devices become more powerful, they may serve as an efficient interface to make up visual, memory, and other deficiencies due to aging • The society is aging – For example, people of 65 and older are 16.8% of Florida’s population (US Census Bureau, 2005) • By modifying and enhancing environments, vision technology can be critical for helping people stay active and be independent 10/12/2008 9:52:30 PM intro-seminar-08.ppt Motivations The goal is to organize large number of images so that they can be retrieved efficiently and effectively based on content 10/12/2008 9:52:51 PM intro-seminar-08.ppt Problems And Approach Organization of large libraries of images for efficient content-based indexing and retrieval of images Image Categorization - We assume that a training database containing labeled images representing various different classes is available • The goal is to learn optimal low-dimensional features that can be used to assign a new query image to the correct class Retrieval - The objective is to find the top ℓ matches in a database to a query image Image 10/12/2008 9:52:53 PM intro-seminar-08.ppt Preliminary Experiments With SH-features Dataset: COREL-1000 (100 images in each of 10 categories) We utilize a bank of 5 filters and apply each filter to the R, G, and B channels of the images to obtain a total of fifteen 11-bin histograms per image • Thus, the SH-feature vector h(I, F) has dimension 165 For a query image I, we calculate the Euclidean distances between h(I, F) and h(J, F), for every J in the database, and rank the images according to increasing distances • To quantify retrieval performance and compare the results with those reported for other methods, we use the weighted precision and the average rank 10/12/2008 9:52:55 PM intro-seminar-08.ppt The Weighted Precision For a query image I, the retrieval precision for the top ℓ returns is nℓ(I)/ℓ, where nℓ is the number of correct returns The weighted precision for COREL-1000 for I is defined as 10/12/2008 9:52:55 PM intro-seminar-08.ppt The Average Rank For each query image, we rank all the images in the dataset • The average rank is the average rank of all images that belong to the same class as the query image • For COREL-1000, the perfect value for a query images is 50.5 10/12/2008 9:52:56 PM intro-seminar-08.ppt Average Precision: Comparison 10/12/2008 9:52:56 PM intro-seminar-08.ppt Average Rank: Comparison 10/12/2008 9:52:57 PM intro-seminar-08.ppt Content-based Image Retrieval We now apply the image-categorization classifier learned with OFA to retrieve images • Note that the low-dimensional image representation was optimized to categorize images correctly according to the nearest neighbor criterion, but not to rank matches to a query image correctly according to their distances in feature space The goal is to exploit the image categorization method to retrieve images • Idea is to use the distance in feature space to assign probabilities that measure the compatibility of the image with a given class. Use these probabilities to rank classes and retrieve images using this ranking 10/12/2008 9:52:57 PM intro-seminar-08.ppt Image Retrieval - continued 10/12/2008 9:52:58 PM intro-seminar-08.ppt Image Retrieval - continued a query image I and a positive integer ℓ, the goal is to retrieve a ranked list of ℓ images from the database. Given • We assume that all images in the database have been indexed according to content using the classifier learned with OFA. • Rank the classes according to the probabilities p(i|I). • Select as many images as possible from the most likely class. • Within this class, images are retrieved and ranked according to their Euclidean distances to I as measured in the reduced feature space (several variants possible). • Once that class is exhausted, we proceed similarly with the second most likely class and iterate the procedure until ℓ images are obtained. 10/12/2008 9:52:59 PM intro-seminar-08.ppt Experimental Results OFA was used to “learn” a 9-dimensional linear reduction of the SH-features of dimension 165, with 400 training images, 40 from each class • Leave-one-out was used for cross-validation The entire database was indexed with the nearest neighbor classifier applied to the reduced features All 1,000 images were used as query images • For each class i, the average weighted precision and the average rank were calculated for comparison with SIMPLIcity and Spectral Histograms 10/12/2008 9:52:59 PM intro-seminar-08.ppt Average Precision: Comparison 10/12/2008 9:53:00 PM intro-seminar-08.ppt Average Rank: Comparison 10/12/2008 9:53:11 PM intro-seminar-08.ppt Examples The top-left image is the query image, which is also the top return 10/12/2008 9:53:12 PM intro-seminar-08.ppt More Examples 10/12/2008 9:53:13 PM intro-seminar-08.ppt Precision–Recall 10/12/2008 9:53:14 PM intro-seminar-08.ppt Average Recall vs. Average Precision 10/12/2008 9:53:14 PM intro-seminar-08.ppt Average Recall vs. Average Precision 10/12/2008 9:53:15 PM intro-seminar-08.ppt Invariant Content Characterization Based on Single View 3D Reconstruction Many of the images share the following characteristics • There is a ground plane, which is relative flat • On the ground plane, we have objects that can be approximated by planar sides – Note the images will change depending on the view angles 10/12/2008 9:53:15 PM intro-seminar-08.ppt Single View 3D Reconstruction We want to characterize the meaningful contents of images • Ground itself in general is not interesting • We want to “capture” the objects on the ground 10/12/2008 9:53:17 PM intro-seminar-08.ppt Ground Plane Estimation Using Horizon Line For a pinhole camera model, each plane at the infinity will become a line in the image 10/12/2008 9:53:18 PM intro-seminar-08.ppt Pinhole Camera Model 10/12/2008 9:53:18 PM intro-seminar-08.ppt Plane at Infinity AX BY CZ D 0 Axcam Z By cam Z CZ D 0 f f Axcam By cam D C 0 f f Z Axcam By cam C 0 if Z f f That is, we can estimate the ground plane’s normal if we can estimate the horizon in the image 10/12/2008 9:53:46 PM intro-seminar-08.ppt Vertical Planes After we estimate the ground plane, we can then estimate in 3D any vertical plane if we know (at least) two pixels on the intersection line between the vertical plane and ground plane • We now approximate the scene using boxes in 3D 10/12/2008 9:53:49 PM intro-seminar-08.ppt Invariant Representations of Content For content-based retrieval, we derive a representation of an image that is scale and view independent • By representing each side of each 3D box using a normalized view – In other words, for each side of the box, we place a virtual camera at a fixed distance (which fixes the scale) and a fixed view, whose up vector is perpendicular to the ground plane (which fixes the view) 10/12/2008 9:53:50 PM intro-seminar-08.ppt An Example Shape Theory We want to quantify the difference between two shapes in a principled way • We do this by constructing a shape space and then use the geodesic distance of two shapes on the shape manifold as the metric 10/12/2008 9:53:54 PM intro-seminar-08.ppt Shape Clustering 10/12/2008 9:53:56 PM intro-seminar-08.ppt Shape Clustering 10/12/2008 9:53:57 PM intro-seminar-08.ppt Clustering Dendrogram 10/12/2008 9:53:57 PM intro-seminar-08.ppt Sulcal Curves Sulcal curves are important for characterizing brain functions 10/12/2008 9:53:58 PM intro-seminar-08.ppt Sulcal Curves Sulcal curves are important for characterizing brain functions 10/12/2008 9:54:00 PM intro-seminar-08.ppt Clustering of Sulcal Curves 10/12/2008 9:54:01 PM intro-seminar-08.ppt Modeling Mathematical Abilities and Disabilities As it is possible to acquire detailed surfaces of the human brain, one may ask how characteristics of the brain structure affect the mathematical abilities and disabilities • The U.S. Department of Education wants to know so that they can understand and find solutions to the mathematical problems young children have Corpus callosum examples of young children without mathematical disabilities (a) and with (b) 10/12/2008 9:54:02 PM intro-seminar-08.ppt Surface Parametrization 10/12/2008 9:54:17 PM intro-seminar-08.ppt Geodesic Interpolation Between Surfaces 10/12/2008 9:54:21 PM intro-seminar-08.ppt Atlas for Hippocampus 10/12/2008 9:54:23 PM intro-seminar-08.ppt Model for Blindness 10/12/2008 9:54:26 PM intro-seminar-08.ppt Automated Alignment of Large Surfaces 10/12/2008 9:54:28 PM intro-seminar-08.ppt Robust Visual Inference With a common domain for surface representations, we can pose the visual inference in the Bayesian framework by building probability models 10/12/2008 9:54:31 PM intro-seminar-08.ppt Human-Robot Collaborative Interaction The goal is to let robots be aware of the positions, poses, expressions, moods, and other factors of the humans so that robots can interact with humans collaborative In collaboration with Prof. Emmanuel Collins at the College Engineering 10/12/2008 9:54:35 PM intro-seminar-08.ppt Automated 3D Phenotype Measurement The central problem in biology is to understand the relationship between genotype and phenotype • With availability of genomes of humans and model organisms, the central problem becomes how to measure phenotype at a large scale 10/12/2008 9:54:39 PM intro-seminar-08.ppt Computer Vision for Computational Systems Biology The goal of systems biology is to link the molecular and cellular events and properties to physiological functions 10/12/2008 9:54:42 PM proteins to organs: The Physiomeintro-seminar-08.ppt Source: “Integration from Project”, Nature Review, Vol. 4, 2007. Live Cell Imaging at Cellular Level 10/12/2008 9:58:24 PM intro-seminar-08.ppt MRI / fMRI / CT / PET Imaging 10/12/2008 9:58:39 PM intro-seminar-08.ppt FISHFinder@FSU Source: Gilbert’s group at Biology Department, FSU 10/12/2008 10:00:00 PM intro-seminar-08.ppt 134 Courses Most Relevant Courses • • • • • • CAP 5638 Pattern Recognition – Offered Spring 2008 CAP 5415 Principles and Algorithms of Computer Vision CAP 6417 Theoretical Foundations of Computer Vision STA 5106 Computational Methods in Statistics I STA 5107 Computational Methods in Statistics I I ISC 5935-05/STA 5934-01 Applied Machine Learning – Offered Spring 2008 • Seminars and advanced studies Related Courses • CAP 5615 Artificial Neural Networks • CAP 5600 Artificial Intelligence 10/12/2008 10:01:07 PM intro-seminar-08.ppt Funding of the Group National • • • • Science Foundation DMS CISE IIS ACT CCF National 10/12/2008 10:01:09 PM Institute of Health intro-seminar-08.ppt Summary Computer Vision Group offers interesting research topics/projects • Efficient represents for generic images and videos • Real-time detection and recognition of objects • Computational models for object recognition and image classification • Medical/biological image analysis • Motion/video sequence analysis and modeling • They are challenging, interesting, and exciting • Now it is a productive and fruitful area to be in 10/12/2008 10:01:16 PM intro-seminar-08.ppt Contact Information • • • • • Name Web sites Email Offices Phones 10/12/2008 10:01:23 PM Xiuwen Liu http://cavis.fsu.edu http://fsvision.fsu.edu http://www.cs.fsu.edu/~liux liux@cs.fsu.edu LOV 166 and 118 North Woodward Ave. 644-0050 and 645-2257 intro-seminar-08.ppt Thank you! Any questions?