Home
Fabio de Miranda's Library tagged → View Popular
PLoS Computational Biology: Why is Real-World Visual Object Recognition Hard?
Progress in understanding the brain mechanisms underlying vision requires the construction of computational models that not only emulate the brain's anatomy and physiology, but ultimately match its performance on visual tasks. In recent years, “natural” images have become popular in the study of vision and have been used to show apparently impressive progress in building such models. Here, we challenge the use of uncontrolled “natural” images in guiding that progress. In particular, we show that a simple V1-like model—a neuroscientist's “null” model, which should perform poorly at real-world visual object recognition tasks—outperforms state-of-the-art object recognition systems (biologically inspired and otherwise) on a standard, ostensibly natural image recognition test. As a counterpoint, we designed a “simpler” recognition test to better span the real-world variation in object pose, position, and scale, and we show that this test correctly exposes the inadequacy of the V1-like model. Taken together, these results demonstrate that tests based on uncontrolled natural images can be seriously misleading, potentially guiding progress in the wrong direction. Instead, we reexamine what it means for images to be natural and argue for a renewed focus on the core problem of object recognition—real-world image variation.
robots.net - Robust Optical Flow Model Derived From Biology
Brinkworth and O'Carroll have come up with a new optical flow model, described in their paper, Robust Models for Optic Flow Coding in Natural Scenes Inspired by Insect Biology (PDF format). Their new system is based on the actual neural processing pathways of the fly which could prove to be a very robust velocity estimator and accurate sensor for self-motion in robots. CC-licensed image of female Tabanus Horse Fly by flickr user Thomas Shahan.
The Art of Jim Campbell: Seeing In Pixels - Boing Boing
Movement makes up for the lack of other visual information. Your brain can read and understand a video at much lower resolution than it would need to make equal sense of a still frame.
LPP - J. Kevin O'Regan
Papers about sensorimotor loop, situated perception, situated vision, enactive vision, visual contingencies
The evolution of object categorization and the challenge of image abstraction Sven Dickinson
PDF: Analyzing Vision at the Computational Level. J. K. Tsotsos
Excellent paper trying to apply computational complexity to analyze some visual tasks and a deep discussion following it.
Review of Visual Tracking
This report contains a review of visual tracking in monocular video sequences. For the purpose of this review, the majority of the visual trackers in the literature are divided into three tracking categories: discrete feature trackers, contour trackers, and region-based trackers. This categorization was performed based on the features used and the algorithms employed by the various visual trackers. The first class of trackers represents targets as discrete features (e.g. points, sets of points, lines) and performs data association using a distance metric that accommodates the particular feature.
Situated Vision in a Dynamic World: Chasing Objects
We describe a system that approaches and follows arbitrary moving objects in real time using vision as its only sense. The system uses multiple simple vision computations which, although individually unreliable, complement each other in a manner mediated by a situated control network.
Visual Attention Research: The FINST Visual Indexing Theory
Work related to Aaron Sloman ideas of registration along the optic array and also situated vision.
Indoor Scene Recognition, CVPR 09
Indoor scene recognition is a challenging open problem in high level vision. Most scene recognition models that work well for outdoor scenes perform poorly in the indoor domain. The main difficulty is that while some indoor scenes (e.g. corridors) can be well characterized by global spatial properties, others (e.g., bookstores) are better characterized by the objects they contain. More generally, to address the indoor scenes recognition problem we need a model that can exploit local and global discriminative information.
Keith Price Bibliography Integration of Vision Modules
Some papers on organisation of (artificial) visual systems
Dynamical Systems, Control, Coding, Computer Vision New Trends,Interfaces, and Interplay
This is just a mention, not the book.
Interesting bit: "As J. Malik and P. Perona point out in "Introduction to Mathematical Aspects of Computer Vision", the vision tasks may be organized in four broad categories: reconstruction of images, control (e.g. visually guided control of vehicle navigation), grouping/tracking and recognition."
Ullman's Visual Routines x CMU Tekkotsu's Sketches
Describes some of the resources available in CMU Tekkotsu robotics simulator and how they relate to Ullman's Visual Routines
To afford or not to afford: A new formalization of affordance towards affordance-based robot control
The affordances of the environment are what it offers the animal, what it provides or furnishes, either for good or ill. The verb to afford is found in the dictionary, but the noun affordance is not. I have made it up. I mean by it something that refers to both the environment and the animal in a way that no existing term does. It implies the complementarity of the animal and the environment.”“... an affordance is neither an objective property nor a subjective property; or both if you like. An affordance cuts across the dichotomy of subjective-objective and helps us to understand its inadequacy. It is equally a fact of the environment and a fact of behavior. It is both physical and psychical, yet neither. An affordance points bothways, to the environment and to the observer.”“The perceiving of an affordance is not a process of perceiving a value-free physical object to which meaning is somehow added in a way that no one has been able to agree upon; it is a process of perceiving a value-rich ecological object.”“The theory of affordances rescues us from the philosophical muddle of assumingfixed classes of objects, each defined by its common features and then given a name...You do not have to classify and label things in order to perceive what they afford.”
Tekkotsu: Cognitive Robotics
NOTE: their definition of cognitive robotics is a little off
... extending Tekkotsu to support higher level abstractions for robot programming. We call our approach "cognitive robotics" because it draws inspiration from ideas in cognitive science, such as Ullman's notion of "visual routines", Paivio's "dual coding theory", and Gibson's concept of "affordances". Cognitive robotics makes no strong theoretical claims, but instead focuses on developing technology that actually works, on real robots, with their inherent physical constraints and sensory limitations.
MeMoMan — Markerless Tracking from Videos
We are developing new computational models and a system for accurate measurement of human motion. Our primary goal is to develop markerless vision-based tracking algorithms for use with the industry-proven anthropometric human model RAMSIS (in collaboration with the TUM Ergonomics Department/Faculty of Mechanical Engineering). By providing RAMSIS with markerless tracking capabilities, we open up new fields of application in ergonomic studies and industrial design. On the other hand, we believe that a far-developed, flexible and accurate model such as RAMSIS is beneficial for human motion tracking given the ergonomic expertise that has affected its design.
International Symposium on Visual Computing
The purpose of the International Symposium on Visual Computing (ISVC) is to provide a common forum for researchers, scientists, engineers and practitioners throughout the world to present their latest research findings, ideas, developments and applications in the broader area of visual computing.
ISCV seeks papers describing contributions to the state of the art and state of the practice in the broad field of visual computing. In particular, ISVC is structured around four central areas of visual computing: (1) computer vision, (2) computer graphics, (3) virtual reality, and (4) visualization. In particular, we are interested in papers that combine technologies from two or more of these areas.
This is the fith in the series of symposia following four successful meetings in 2005, 2006, 2007 and 2008. ISVC09 will consist of invited and contributed presentations dealing with all aspects of visual computing. In addition to the main symposium, the symposium will include six keynote presentations, several special tracks, and a poster session. Also, ISVC09 is hosting the 3rd Semantic Robot Vision Challenge. Two, $500 "best paper awards" are sponsored by Mitsubishi Electric Research Labs and by Volt / Microsoft MSDN. Also, Informatics Circle of Research Excellence (iCORE) sponsors a $500 award for the best paper in the special track on Visualization Enhanced Data Analysis for Health Applications.
[Imageworld] CFP: 3rd Semantic Robot Vision Challenge 2009
The Semantic Robot Vision Challenge (SRVC) is a research competition
that is designed to advance the state of the art in embodied vision,
active scene understanding, and automatic acquisition of knowledge
from the Internet. This competition is an indoor robotic photo
scavenger hunt. Teams will be required to demonstrate a robot that has
the ability to:
1. Read and parse a textual list of generic or specific objects to be
found in the environment. Examples from 2008 include “frying pan”,
“upright vacuum cleaner” and “Ritz Crackers”.
2. Autonomously connect to the Internet and build an object
classification database from images or other information related to
the listed objects, in a set amount of time.
3. Autonomously navigate and search the unknown environment with the
task of taking snapshots of the listed objects using the data acquired
from the Internet.
4. Return an image of each object type containing a single bounding
box around the pictured object.
Imageworld Info Page
Imageworld is used to announce worldwide events and academic vacancies within the field of Computer Vision, Image Analysis, and Medical Image Analysis.
Selected Tags
Related Tags
Sponsored Links
Top Contributors
Groups interested in computer...
Diigo is about better ways to research, share and collaborate on information. Learn more »
Join Diigo