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.
Brightkite
Annotate the world with a mobile device
foursquare
Geocoded / geolocated mobile app for real-world tagging
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.
6 free remote desktop sharing and screen sharing solutions
6 free remote desktop sharing and screen sharing solutions
The dark side of Dubai (The Independent, UK)
Dubai was meant to be a Middle-Eastern Shangri-La, a glittering monument to Arab enterprise and western capitalism. But as hard times arrive in the city state that rose from the desert sands, an uglier story is emerging.
sndpeek : real-time audio visualization
sndpeek is just what it sounds (and looks) like:
real-time 3D animated display/playback
can use mic-input or wav/aiff/snd/raw/mat file (with playback)
time-domain waveform
FFT magnitude spectrum
3D waterfall plot
lissajous! (interchannel correlation)
rotatable and scalable display
freeze frame! (for didactic purposes)
real-time spectral feature extraction (centroid, rms, flux, rolloff)
available on MacOS X, Linux, and Windows under GPL
part of the sndtools distribution.
digitaltutor.net - About iList - tool to learn linked lists
iList is an Intelligent Tutoring System that can help you learn linked lists. It allows you to work on linked lists problems within an interactive, user friendly environment. However, iList is more than just a graphical environment: iList carefully monitors your progress and gives you helpful feedback while you are working towards a solution. The system does so by using sophisticated artificial intelligence technology. If you want to learn more about the science behind iList, you will find plenty of information in the papers referenced in the documentation section.
Jeroo! Programming learning tool
Jeroo is an effective, award winning, classroom-tested tool that helps novices learn fundamental concepts of object-oriented programming, including
Instantiating and using objects
Writing methods to extend behavior
Selecting and using fundamental control structures
Jeroo engages students with
Story telling
Animated execution
Simultaneous code highlighting
Java Binding for the OpenCL API: Wiki: Home — Project Kenai
This project provides a easy to use Java binding for the OpenCL API and is released under the BSD license. GlueGen is used to generate a low level binding directly from the official Khronos C header files. A hand written high level binding on top of generated code provides a convenient interface and reduces verbosity to a minimum.
Startups.com | Your Business. Your Questions.
Stackexchange site, somewhat duplicates answers.onstartups.com
3D modelling with webcam - nice papers and videos
The generation of 3D models is very useful for many computer vision applications. This paper introduces ProFORMA, a system designed to enable on-line reconstruction of textured 3D objects rotated by a user's hand. Partial models are created very rapidly and displayed to the user to aid view planning, as well as used by the system to robustly track the object pose. The system works by calculating the Delaunay tetrahedralisation of a point cloud obtained from on-line structure from motion estimation which is then carved using a recursive and probabilistic algorithm to rapidly obtain the surface mesh.
tmux
tmux is a terminal multiplexer: it enables a number of terminals (or windows), each running a separate program, to be created, accessed, and controlled from a single screen. tmux may be detached from a screen and continue running in the background, then later reattached.
tmux uses a client-server model. The server holds multiple sessions and each window is a independent entity which may be freely linked to multiple sessions, moved between sessions and otherwise manipulated. Each session may be attached to (display and accept keyboard input from) multiple clients.
appscale - Project Hosting on Google Code
AppScale is a platform that allows users to deploy and host their own Google App Engine applications. It executes automatically over Amazon EC2 and Eucalyptus as well as Xen and KVM. It has been developed and is maintained by the RACELab at UC Santa Barbara.
typhoonae - Project Hosting on Google Code
The TyphoonAE project (formerly known as twistedae) aims at providing a full-featured and productive serving environment to run Google App Engine (Python) applications. It delivers the parts for building your own scalable App Engine while staying compatible with Google's API.
AEJ Tools - App Engine REST Tools
Look at this line of code:
datastore.query(Select.from("Movie").where("title").is("Star Wars"));
Jython runs on Google AppEngine! at Jython Journeys
If you’re not yet aware of it, Google AppEngine is one of the foremost Cloud Computing offerings currently available. It is also arguably one of the purest options available, since it truly removes the need for the application administrator to consider physical resources (apart from paying for them that is!). The Google AppEngine cloud will automatically scale up the resources as and when required.
Structured Procrastination
Procrastinators often follow exactly the wrong tack. They try to minimize their commitments, assuming that if they have only a few things to do, they will quit procrastinating and get them done. But this goes contrary to the basic nature of the procrastinator and destroys his most important source of motivation. The few tasks on his list will be by definition the most important, and the only way to avoid doing them will be to do nothing. This is a way to become a couch potato, not an effective human being.
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.
Sponsored Links
Top Tags
- 746programming,
- 381computerVision,
- 363java,
- 273python,
- 245fun,
- 220vision,
- 1903d,
- 186robotics,
- 155computerGraphics,
- 137imageProcessing,
View All Recent Tags (50)
- 34computerVision,
- 30vision,
- 15programming,
- 12business,
- 10java,
- 9python,
- 9robotics,
- 8augmentedReality,
- 6cognitiveScience,
- 6startup,
- 5robots,
- 4computation,
- 4computing,
- 4video,
- 3appengine,
- 3perception,
- 3artificialIntelligence,
- 3visualTasks,
- 3computerGraphics,
- 3jmf,
- 2objectRecognition,
- 2bioinspired,
- 2mobile,
- 2geolocation,
- 2cloudcomputing,
- 2game,
- 2diigo,
- 2neuroScience,
- 2gamedev,
- 2startups,
- 2complexity,
- 2softwareEngineering,
- 2superresolution,
- 2artoolkit,
- 2cognition,
- 2activeVision,
- 2agile,
- 2patternRecognition,
- 2research,
- 2datasets,
- 2via:csantos,
- 2parallel,
- 2concurrent,
- 2imageProcessing,
- 2webdev,
- 2mapping,
- 2iphone,
- 2media,
- 1opticalFlow,
- 1education
Public Tags (1463)
Fabio de Miranda's Public Lists (2)
Highlighter, Sticky notes, Tagging, Groups and Network: integrated suite dramatically boosting research productivity. Learn more »
Join Diigo