Text Only
Asian Conference on Computer Vision,ACCV`98

Final Program

Conference Organization

Welcome Message

Message from Program Co-Chairs

Location Map of HKUST Campus

The Mass Transit Railway of HK (MTR)


Catering Services on Campus

Technical Program

Technical Sessions

Back to Home







To Top of the Page

Welcome Message

The Asian Conference on Computer Vision celebrates its third anniversary in 1998. For the first time, it will be held in Hong Kong, which is now part of the Peoples' Republic of China. The theme of ACCV'98 is "Computer Vision in the Information Era". The conference will provide a forum for researchers to exchange up-to-date technical knowledge and experience. The conference will focus on theory as well as applications of computer vision. You will find 78 papers, grouped into 9 sessions, and 108 poster presentations in our program. Our authors come from 32 countries.

The ACCV'98 Organizing Committee is looking forward to your participation at what promises to be a most enriching event. If you have not yet registered, please do so. Conference and hotel registration information can be found at this advanced program handbook.

As co-chairman of ACCV'98, I would like to thank all members of the Organizing Committee and especially the Program Committee who have worked extremely hard to ensure the success of the conference. Lastly, I would like to express my appreciation to all the public and private agencies who have supported this conference.

Helen Shen
General Co-Chair of ACCV'98

To Top of the Page

Message from Program Co-Chairs

We are very pleased to have the opportunity to organize the 3rd Asian Conference on Computer Vision (ACCV'98). The conference is sponsored by the IEEE Hong Kong Section, Computer Chapter, the Sino Software Research Institute and the Department of Computer Science of the Hong Kong University of Science and Technology, and the Hong Kong Industry Department.

We received over 300 submissions of full papers (not including the invited papers for the special sessions) from 30 countries in April 1997. In order to provide a quality conference and quality proceedings, each paper was reviewed by at least three members of the program committee. The program committee selected and accepted 58 papers for oral presentation and 112 papers for poster presentation after the review process. Some of these papers were jointly submitted to ACCV'98 and ICCV'98 (to be held in Bombay in January 4-7, 1998) and they were reviewed in a coordinated effort. We must add that the program committee and the reviewers have done an excellent job within a tight schedule and we are very pleased with the quality of the papers.

Four eminent invited speakers, Professors Brian Funt of Simon Fraser University, Krishna Nathan of IBM, Eric Grimson of MIT, and Shoji Tominagaof Osaka Electro-Communication University, have contributed to the conference. We are grateful to them. In addition, we wish to thank Professors Jake Aggarwal, Shashi Buluswar, Yi-Ping Hung, Anil Jain, and Sharatchandra Pankanti for organizing the very high-quality special sessions. Last but not least, we would like to express our gratitude to all the contributors, reviewers, program committee and organizing committee members, and sponsors, without whom the conference would not have been possible.

Roland Chin
Ting-Chuen Pong

To Top of the Page

Location Map of HKUST Campus

Location Map of HKUST Map

To Top of the Page

The Mass Transit Railway of HK (MTR)

The Mass Transit Railway of HK (MTR)

To Top of the Page


You can come to the campus by one of the following ways:
1. By Taxi
The following is the address of the HKUST campus in Chinese. If necessary, show it to the taxi driver.
HKUST Address in Chinese

2. By Public Transportation
To be adventurous, you can take the MTR (route map) and then a bus.

From the Choi Hung MTR Station, Exit "B", at the bus stop by the street, you can take either the No. 91 or 91M (double decker) bus or the No. 11 (green van) mini-bus

From the Lam Tin MTR station, Exit "B", you can take the No. 298 bus at the bus terminal.

To Top of the Page

Catering Services on Campus

LocationServiceTypes of Food
Academic Concurse Coffee Shop Western, Sandwiches, coffee, desert
LG5 Food Court Chinese, South East Asian, Japanese, Wendy's
LG7 Caferteria Mainly Chinese
University Center Restaurant Western, Italian

To Top of the Page

Technical Program

Thursday, Jan. 8 Friday, Jan. 9 Saturday, Jan. 10
9:00am-10:15am Invited talk:
   Krishna Nathan
Invited talk:
    Eric Grimson
Invited talks:
    Brian Funt and Shoji Tominaga
10:15am-10:30am Coffee break Coffee break Coffee break
10:30am-12:30pm Special Session T1A:
    Biometry I

Session T1B:
    Physics-Based Vision
Special Session F1A:
    Biometry II

Session F1B:
    Low-level processing
Special Session S1A:
    Recent Advances in CV

Session S1B:
    Grouping and Segmentation
12:30pm-1:30pm Lunch Lunch Lunch
1:30pm-3:30pm Special Session T2A:
    Color Vision I

Session T2B:
    Robot Vision and Navigation
Special Session F2A:
    Color Vision II

Session F2B:
    Active Vision
Special Session S2A:
    Computer Vision & Virtual Reality

Session S2B:
    Motion Analysis
3:30pm-3:45pm Coffee break Coffee break Coffee break
3:45pm-5:45pm Session T3A:
    OCR and Applications

Poster Session I
Session F3A:
    Face and Hand Posture Recognition

Poster Session II
Session S3A:
    Object Recognition and Modeling

Poster Session III
6:00pm Reception Banquet

* Invited Talks will be held in LTB
* Session xxA will be held in LTB
* Session xxB will be held in LTC
* All Poster Sessions will be held at the Academic Concourse
* Coffee Breaks will be held at the Academic Concourse

* Room 1403 is assigned to be the authorsí preparation room.

* Room 1401, 1402 are conference offices for general enquiry

Thursday, January 8, 1998.

Invited Talk:

Pen Computing -- An Overview

Krishna S. Nathan
IBM T. J. Watson Research Center, U.S.A.

Pen computing has had its share of ups and downs over the past few years - from the heady promises of handwriting recognition and Personal Digital Assistants (PDAs) to the disenchantment due to poor sales of pen based computers and the lack of credible applications for them. In my talk, I will discuss the issues that have led to this state of affairs, as well the current trends in the industry. The main challenges facing pen computing today will also be explored. These include improving algorithms for handwriting recognition, building better and more natural hardware platforms and designing user interfaces that take into account the demands multimodal input. The talk will also touch upon work in this area that is being carried out at the T.J. Watson Research Center at IBM.

Speakerís Biography:
Krishna S. Nathan is Senior Manager of Consumer Devices at IBM T.J. Watson Research Center in Yorktown Heights, NY where he is responsible for research and development activities in handwriting recognition, novel mobile pen platforms and related applications.

Dr. Nathan joined IBM Research in 1991 after receiving a doctoral degree in Electrical Sciences from Brown University in the area of speech recognition. Prior to that he was a Member of Technical Staff at the Jet Propulsion Laboratory at the California Institute of Technology. He holds a master degree in Electrical Engineering from the Massachusetts Institute of Technology in Cambridge, Ma. His professional interests are in the areas of pattern recognition, signal processing and mobile computing.

Friday, January 9, 1998.

Invited Talk:

Image Guided Surgical Systems

Eric Grimson M.I.T., U.S.A.

Imagine giving a surgeon an "X-Ray vision" ability, that is, allowing him to look at a patient and see through skin, bone, fat to visualize internal structures, like tumor or blood vessels, exactly as they occur within the patient. Imagine allowing him to insert instruments through narrow openings in the body, yet still be able to visualize the full anatomical context around the instrument. Such capabilities would enable a surgeon to better plan procedures, to navigate through delicate procedures and to evaluate the effectiveness of surgical procedures, e.g "have I removed all of the tumor". This talk will describe such an image guided surgical system, currently in regular use at Brigham and Women's Hospital in Boston for neurosurgical cases. The system combines a suite of computer vision techniques, and allows a surgeon to visualize internal structures in registration with the patient, to interactively probe points within the patient and see the full anatomtical context of his position, and to evaluate progress of the surgical procedure. Related techniques allow a clinician to noninvasively map functional areas of the cortex, such as motor cortex, in order to provide the surgeon with registered information about critical structures during surgery.

Speakerís Biography:

Eric Grimson received a B.Sc. (Hons) in Mathematics and Physics from the University of Regina in 1975 and a Ph.D. in Mathematics from MIT in 1980. He is currently a Professor of Computer Science and Engineering at MIT and a Lecturer on Radiology at Harvard Medical School. For the past twenty years, he has conducted research on computer vision, and has authored two technical monographs in this field. Prof. Grimson currently heads the Computer Vision Group of MIT's Artificial Intelligence Laboratory, which has pioneered state of the art systems for object recognition, image database indexing, image guided surgery, target recognition, site modeling and many other areas of computer vision. Recently, his group has been active in applying vision techniques in medicine: for image guided surgery minimally invasive surgery and telemedicine.

Saturday, January 10, 1998.

Invited Talk:

Computational Color: Digital Photography and Computer Vision

Brian Funt
Simon Fraser University, B.C., Canada

The members of the Computational Vision Laboratory at Simon Fraser University have been studying color for over a decade. I will discuss some of the main color issues, the progress we have made in the understanding them and the application of our methods to color-based object recognition and digital photography.

To explain color perception, we much explain how it is that we see colors as relatively stable despite changes in the incident illumination. I make the assumption that color like the rest of visual perception is there to give us information about the world, the surface properties of objects in particular, and so the stability and reliability of the information is important. The problem of color stability arises because the light reaching our eyes from an object is the product of the objectís surface reflectance and the spectrum of the light illuminating the object. We do no have direct access to the properties of the incident light, so somehow we must estimate them from the light we receive from the object. To make matters worse, our eyes only measure the spectrum at extremely low resolution.

Clearly, a stable representation of object color would be useful in computer vision. It is equally important for digital photography since we do not want images to produce images whose color balance depends on the qualities of the ambient scene illumination. Until recently the growth of digital photography was limited less by camera technology than by the lack of economical digital printers with photographic quality a situation that has changed gramatically in the last six months with the introduction of some new ink jet printer technology (e.g. the Hewlett-Packard Photo Smart Printer). Digital photography is providing new impetus to the search for better model of color perception. Comparative tests of many different color constancy methods show that two work most reliably: one based on a neural network to estimate the illumination properties from a color histogram, and a second based on the constraints provided by the image gamut.

Speakerís Biography:

Brian Funt is the Director of the Center for Systems Science and Professor of Computing Science at Simon Fraser University where he has been since 1980. He received his B.Sc., M.Sc. and Ph.D.(1976) degrees from the UBC, Canada. His research has focussed on computational models of color perception for 15 years. At iccv-2, he was awarded the Marr prize for ďcolor from Black and WhiteĒ. He is currently an Associate Editor of IEEE Transactions on Image Processing responsible for papers on color in particular.

Saturday, January 10, 1998.

Invited Talk:

Color Vision and Color Media Processing Research in Asia

Shoji Tominaga
Osaka Electro-Communication University, Japan

I will survey the state of color vision and color media processing research in Asia from a computation viewpoint. First, I will discuss briefly the principles of color vision and color measurement. Then I describe why color constancy is important and difficult to be realized in machine vision. The goal of computational color constancy is to recover the physical properties of illuminant and surface from photosensor responses. I introduce a vision system for color constancy and our current results.

Next, I will discuss algorithms for understanding complex color images. Real images exhibit rich and complex structure, whose nature is determined by the physical and geometric properties of illumination, reflection, and imaging. I suggest some reflection models adequate for describing light reflection of a variety of materials. Then approaches are introduced for estimating scene parameters based on the reflection models.

Finally, I will discuss the issue of color management. The concept of device-independent color reproduction of image has received wide spread attention since the advent of desk-top publishing systems. Color reproduction requires color conversion between the color signals, depending on a device, and the standard color coordinates, representing color appearance. Mapping methods are introduced for solving the color conversion problem for color printers.

Speakerís Biography:

Shoji Tominaga was born in Hyogo Prefecture, Japan, on April 12, 1947. He received the B.E., M.S., and Ph.D. degrees in electrical engineering from Osaka University, Toyonaka, Osaka, Japan, in 1970, 1972, and 1975, respectively. From 1975 to 1976, he was with Electrotechnical Laboratory. Since April 1976, he has been with Osaka Electro-Communication University, Neyagawa, Osaka, where he is currently a Professor with the Department of Engineering Informatics. During the 1987-1988 academic year he was a Visiting Scholar at the Department of Psychology, Stanford University. His research interests include computational color vision, color image analysis, and neural networks. He is a member of the Optical Society of America, IEEE, SID and IS&T.

To Top of the Page

Conference Programme: Technical Sessions

Thursday, January 8, 1998

9:00am - 10:15am Invited Talk

Pen computing - an overview

10:15am - 10:30am Coffee Break

10:30am - 12:30pm Session T1A and Session T1B

Session T1A: Special Session on Biometry I

1. Research Issues in Biometrics*
2. Automatic On-line Signature Verification*
3. Integrating Faces and Fingerprints for Personal Identification
4. Automated Fingerprint Pattern Classification Error Analysis*
5. A High-Dimensional Indexing Scheme for Scalable Fingerprint-Based Identification*

Session T1B: Physics-Based Vision

1. Sign of Surface Curvature from Shading Images Using Neural Network
2. On the Classification of Singular Points for the Global Shape from Shading Problem: A Study of the Constraints Imposed by Isophotes
3. Determination of Sign of Gaussian Curvature of Surface Having General Reflectance Property
4. Estimating Depth Through the Fusion of Photometric Stereo Images
5. Out of the Dark: Using Shadows to Reconstruct 3D Surfaces

12:30pm - 1:30pm Lunch

1:30pm - 3:30pm Session T2A and Session T2B

Session T2A: Special Session on Color Vision I

1. Estimation of Reflection Parameters from a Color Image*
2. Color Edge Detection using Orthogonal Polynomials
3. A Color Normalization Algorithm for Image Indexing*
4. Adaptive Color-Image Embeddings for Database Navigation*
5. A Large Capacity Steganography Using Color BMP Images*

Session T2B: Robot Vision and Navigation

1. Dynamic Calibration of an Active Vision System to Compute the Ground Plane Transformation
2. Identification of 3D Reference Structures for Video-Based Localization
3. Directing Robots with Visual Primitives for Navigation and Micro-manipulation
4. Combining Camera and Laser Radar for ALV Navigation
5. Stereo Vision-Based Obstacle Detection for Partially Sighted People

3:30pm - 3:45pm Coffee break

3:45pm - 5:45pm Session T3A and Session T3B

Session T3A: OCR and Applications

1. Evaluation and Application of Recognition Confidence in OCR
2. A New Nonlinear Shape Normalization Method for Off-line Handwritten Chinese Character Recognition
3. A Novel Triangulation Procedure for Thinning Cursive Text
4. Digital Geometric Methods in Image Analysis and Compression
5. Detection and Enhancement of Small Masses via Precision Multiscale Analysis
6. A Method of Industrial Parts Surface Inspection Based on an Optics Model

Poster Session I

1. Illumination Color from the Blurred Inter-reflection of a Reference Nose
2. Shape Recovery from One Image under Multiple Light Sources
3. Spherical and Cylindrical Light Source Models for Shape Recovery
4. Polyhedral Shape Recovery Based on Interreflections
5. Improved Supervised Color Constancy for Color Inspection
6. Unsupervised Filtering of Munsell Spectra
7. Foveated Vision for Scene Exploration
8. Evolutionary Methods Applied to Binocular Disparity Estimation
9. Robust Epipolar Geometry Estimation Using Genetic Algorithm
10. New Development of Stereo Vision: A Solution of Motion Stereo Correspondence
11. Acquisition of Three-Dimensional Information Using Omnidirectional Stereo Vision
12. Error Analysis in Stereo Vision
13. Detecting Targets in SAR Images: A Machine Learning Approach
14. Precise Matching by Robust Estimation of Deformation and Local Coherence
15. Active Viewpoint Control for Shape from Occluding Contours
16. Point Selection: A New Comparison Scheme for Size Functions (With an Application to Monogram Recognition)
17. Sketch Up: Towards Qualitative Shape Data Management
18. Robust Matching and Hierarchical Recognition of 2-D Shapes Using "Chain of Circles"
19. Finding the Center of Rotational Symmetry from Noisy Forms
20. Recognition in Wavelet-Compressed Imagery
21. Fast Image Template and Dictionary Matching Algorithms
22. Recognition of Planar Shapes Using Algebraic Invariants from Higher Degree Implicit Polynomials
23. Object Recognition and Orientation via Zernike Moments
24. A Study of Zernike Moment Computing
25. Query Expansion by Raw Image Features and Text Annotations in Image Retrieval
26. Montage: An Image Database for the Fashion, Textile, and Clothing Industry in Hong Kong
27. Auto Cameraman Via Collaborative Sensing Agents
28. Dynamic Adaptive Data Structures for Semantic Analysis and Synthesis of Video Information
29. Recognition of Simple Curved Surfaces from 3D Surface Data
30. A Recursive Fitting-and-Splitting Algorithm for 3-D Object Modeling Based on Superquadrics
31. Learning and Recognizing 3D Objects by Using Partial Planar Curve Matching Method
32. Contour Matching Technique for 3D Object Recognition Using Kalman Filter
33. Kalman Filter Based Matching Technique for 3D Object Recognition
34. A Generating Method for 3-dimensional Knitting Cloth Shapes
35. A Fast Mesh Deformation Method to Build Spherical Representation Models of 3D Objects
36. Semi-automatic 3D Object Digitizing System Using Range Images

6:00 - 7:30 pm Reception at the University Center

Friday, January 9, 1998

9:00am - 10:15am Invited Talk

Image Guided Surgical Systems

10:15am - 10:30am Coffee Break

10:30am - 12:30pm Session F1A and Session F1B

Session F1A: Special Session on Biometry II

1. Technical Evaluation of Biometric Systems*
2. Face Recognition from Sequences Using Models of Identity
3. Enhancing Human Face Detection using Motion and Active Contours
4. Learning Identity and Behaviour with Neural Networks
5. Open Sesame! Speech, Password or Key to Secure Your Door?*

Session F1B: Low-Level Processing

1. A Unified Framework for Image-Derived Invariants
2. Stereo Correspondences in Scale Space
3. Fast Stereo Matching in Compressed Video
4. Robust Total least Squares Based Optic Flow Computation
5. Image Processing via the Beltrami Operator

12:30pm - 1:30pm Lunch

1:30pm - 3:30pm Session F2A and Session F2B

Session F2A: Special Session on Color Vision II

1. Efficient Contour Extraction in Color Images*
2. A Natural Norm for Color Processing*
3. Fast and Robust Segmentation of Natural Color Scenes*
4. Segmentation and Tracking Using Color Mixture Models
5. Object Tracking Using Adaptive Color Mixture Models*

Session F2B: Active Vision

1. A Learning Approach to Fixating on 3D Targets with Active Cameras
2. Automatic Detection and Tracking of Human Heads Using an Active Stereo Vision System
3. Front Propagation and Level-Set Approach for Geodesic Active Stereovision
4. A Bayes Nets-Based Prediction/Verification Scheme for Active Visual Reconstruction
5. Actively Building Models with VIRTUE

3:30pm - 3:45pm Coffee Break

3:45pm - 5:45pm Session F3A and Session F3B

Session F3A: Face and Hand Posture Recognition

1. Using RBF Networks to Map GWT Ridge Images to Pose
2. 3-D Pose Estimation and Model Refinement of an Articulated Object from a Monocular Image Sequence
3. Face Synthesis with Arbitrary Pose and Expression from Several Images - An Integration of Image-Based and Model-Based Approaches
4. Live Facial Expression Generation Based on Mixed Reality
5. Real-Time Tracking of Human Hands from a Sign-Language Image Sequence
6. The Model-Based Dynamic Hand Posture Identification Using Genetic Algorithm

Poster Session II

1. Parallel Implementation of Fractal Image Compression Using Multiple Digital Signal Processors
2. Segmentation of MRF Based Image Using Hierarchical Genetic Algorithm
3. Motion Compensated Color Video Classification Using Markov Random Fields
4. Edge-Preserving Smoothing by Convex Minimization
5. On Typical Implementations of Hough Transform for Improving Its Performances
6. Hierarchical Segmentation and Representation with Dynamic Link Architecture Neural Network
7. Perceptually Consistent Segmentation of Texture Using Multiple Channel Filter
8. Optimal Edge Detection under Difficult Imaging Conditions
9. Restoring Image Quality Through Structure Preserving De-noising
10. Feature Saliency from Noise Variations in Invariants
11. Multiscale Image Representation and Edge Detection
12. Rotation Invariant Texture Features from Gabor Filters
13. Euclidean Invariants of Linear Scale-Spaces
14. Segmenting Objects at Multiple Scales : A Robust Approach
15. Multi-grid Edge Models for Magnifying Digital Images
16. Scale and Rotation Invariant Recognition Method Using Higher Order Local Autocorrelation Features of Log-Polar Image
17. Script and Language Identification from Document Images
18. Document Categorization for Document Image Understanding
19. Recognition of Various Bar-graph Structures Based on Layout Model
20. Word-Class Bigram Statistics Language Model for a Hand-Written Chinese Character Recognizer
21. Log Classification by Single X-ray Scans Using Texture Features from Growth Rings
22. Precise and Fast Form Identification Method by Using Adaptive Base Lines for Matching
23. Combinatorial Coarse Classification Method for OLCCR
24. Detecting Characters in Grey-Scale Scene Images
25. Conic Based Image Transfer for 2-D Objects: A Linear Algorithm
26. Minimal Conditions on Intrinsic Paramenters for Euclidean Reconstruction
27. Surface Based Hypothesis Verification in Intensity Images Using Geometric and Appearance Data
28. Next Best Viewpoint (NBV) Planning for Active Object Modeling Based on a Learning-by-Showing Approach
29. Object Recognition by Matching Symbolic Edge Graphs
30. Interpretation of Complex Scenes Using Bayesian Networks
31. Recognition of Urban Scene Using Silhouette of Buildings and City Map Database
32. A Cooperative Inference Mechanism for Extracting Road Information Automatically
33. Model-based Active Object Recognition Using MRF Matching and Sensor Planning
34. Improved Image Classification Using Morphing
35. Reconstruction of Non-manifold Objects from Two Orthographic Views
36. 3D Object Recognition Using Segment-Based Stereo Vision
37. Robust Motion Segmentation Using Rank Ordering Estimators

6:00pm Banquet - The Jumbo Floating Restaurant, Aberdeen, Hong Kong

Saturday, January 10, 1998

9:00am - 10:15am Invited Talk

The State of Color Vision Research
Color Vision and Color Media Processing Research in Asia

10:15am - 10:30am Coffee Break

10:30am - 12:30pm Session S1A and Session S1B

Session S1A: Special Session on Recent Advances in Computer Vision

1. Recent Advances in Detection and Description of Buildings from Multiple Aerial Images*
2. Visual Surveillance of Human Activity*
3. Bayesian Paradigm for Recognition of Objects - Innovative Applications*
4. Toward Motion Picture Grammars*

Session S1B: Segmentation and Grouping

1. Hierarchical Texture Segmentation
2. Range Image Segmentation: Adaptive Grouping of Edges into Regions
3. Optimising the Complete Image Feature Extraction Chain
4. A Unified Framework for Salient Curves, Regions, and Junctions Inference
5. Learning Multiscale Image Models of 2D Object Classes

12:30pm - 1:30pm Lunch

1:30pm - 3:30pm Session S2A and Session S2B

Session S2A: Special Session on Computer Vision & Virtual Reality

1. 3D Model Centered Framework for CV and VR*
2. Image-Based Geometrically-Correct Photorealistic Scene/Object Modeling(IBPhM): A Review*
3. Measuring Object Surface Shape and Reflectance Properties*
4. Robust Image Composition Algorithms for Augmented Reality*
5. Context-Based Recognition of Manipulative Hand Gestures for Human Computer Interaction*

Session S2B: Motion Analysis

1. An Algorithm for Recursive Structure and Motion Recovery under Affine Projection
2. Relative Affine Depth: Structure from Motion by an Uncalibrated Camera
3. The Eigenspace Method for Rigid Motion Recovery from less than Eight Point Correspondences
4. 3D Shape and Motion Analysis from Image Blur and Smear: A Unified Approach
5. 3D Line's Extraction from 2D Spatio-temporal Image Created by Sine Slit
6. Toward Non-intrusive Motion Capture

3:30pm - 3:45pm Coffee break

3:34pm - 5:45pm Session S3A and Session S3B

Session S3A: Object Recognition and Modeling

1. Appearance Based Visual Learning and Object Recognition with Illumination Invariance
2. Evidence-Based Scene Interpretation Considering Subjective Certainty of Recognition
3. Robust Hypothesis Verification for Model Based Object Recognition Using Gaussian Error Model
4. Shape Modeling from Multiple View Images Using GAs
5. 3-D Reconstruction of Multipart Self-Occluding Objects
6. On Analysis of Cloth Drape Range Data

Poster Session III

1. VR Models from Epipolar Images: An Approach to Minimize Errors In Synthesized Images
2. Shape and Pose Parameter Estimation of 3D Multi-Part Objects
3. Generating 3D Models of Objects Using Multiple Visual Cues in Image Sequences
4. Strategical Tracking of Polyhedral Objects by Reactive Change of Projection Pattern - Reactive Range Finder
5. Autonomous Vision-Guided Robot Manipulation Control
6. A New Adaptive Approach on Rapid Obstacle Detection in Range Image
7. Recognition of Shape Model for General Roads
8. Visual Detection of Obstacles Assuming a Locally Planar Ground
9. Potential-Based Modeling of 2D Regions Using Non-uniform Source Distributions
10. A Linear Algorithm for Motion from Three Weak Perspective Images Using Euler Angles
11. On Learning Spatio-Temporal Relational Structures in Two Different Domains
12. An Efficient Iterative Pose Estimation Algirithm
13. A New Multistage Approach to Motion and Structure Estimation by Gradually Enforcing Geometric Constraints
14. Tracking a Person with Pre-recorded Image Database and a Pan, Tilt, and Zoom Camera
15. Recovery of Motion and Structure from Optical Flow under Perspective Projection by Solving Linear Simultaneous Equations
16. Vector Coherence Mapping: A Parallelizable Approach to Image Flow Computation
17. Optical Flow in the Scale Space
18. Motion Detection in Temporal Clutter
19. A Novel Fast Three-Step Search Algorithm for Block-Matching Motion Estimation
20. Moving Vehicle Detection and Tracking in Image Sequences
21. Gesture Recognition from Image Motion Based on Subspace Method and HMM
22. Identifying Faces under Varying Pose Using a Single Example View
23. Multiple Camera Based Human Motion Estimation
24. An Autonomous Facial Caricaturing Based on a Model of Visual Illusion- Experimental Modeling of Visual Illusion
25. 3D Estimation of Facial Muscle Parameter from the 2D Marker Movement using Neural Network
26. Appearance-Based Face Recognition under Large Head Rotations in Depth
27. Skin-Color Modeling and Adaptation
28. Human Information Retrieval by Face Extraction and Recognition on TV News Images Using Subspace Method
29. Converting Facial Expressions Using Recognition-Based Analysis of Image Sequences
30. Muscle-Based Feature Models for Analyzing Facial Expressions
31. A Morphological Method for Moving Object Segmentation and Posture Recognition
32. Detection of Glasses in Facial Images
33. Non-monotonic Continuous Dynamic Programming for Spotting Recognition of Hesitated Gestures from Time-Varying Images
34. Face Recognition Using a Face-Only Database: A New Approach
35. A New Method for Enhancement and Improvement of Image Contrast Based on Lateral Inhibition Model

* Invited papers

To Top of the Page

Back to Home
Back to ACCV'98 Home

Email to ACCV(accv@cs.ust.hk)
Send Comments and Suggestions to accv@cs.ust.hk

To Dept. of CS
To CS Department Homepage

Last Updated: 29th Dec,1997