Montpellier, France, 2000
Artificial Neural Networks have spurred remarkable recent progress in image classification and speech recognition. But even though these are very useful tools based on well-known mathematical methods, we actually understand surprisingly little of why certain models work and others don’t. So let’s take a look at some simple techniques for peeking inside these networks.
We train an artificial neural network by showing it millions of training examples and gradually adjusting the network parameters until it gives the classifications we want. The network typically consists of 10-30 stacked layers of artificial neurons. Each image is fed into the input layer, which then talks to the next layer, until eventually the “output” layer is reached. The network’s “answer” comes from this final output layer.
One of the challenges of neural networks is understanding what exactly goes on at each layer. We know that after training, each layer progressively extracts higher and higher-level features of the image, until the final layer essentially makes a decision on what the image shows. For example, the first layer maybe looks for edges or corners. Intermediate layers interpret the basic features to look for overall shapes or components, like a door or a leaf. The final few layers assemble those into complete interpretations—these neurons activate in response to very complex things such as entire buildings or trees.
One way to visualize what goes on is to turn the network upside down and ask it to enhance an input image in such a way as to elicit a particular interpretation. Say you want to know what sort of image would result in “Banana.” Start with an image full of random noise, then gradually tweak the image towards what the neural net considers a banana (see related work in , , , ). By itself, that doesn’t work very well, but it does if we impose a prior constraint that the image should have similar statistics to natural images, such as neighboring pixels needing to be correlated.
Playscii is an open source ASCII art program, the successor to EDSCII. It runs on Windows and Linux, and will run on Mac OSX soon after a bit more work.
More info: http://vectorpoem.com/playscii/
Please note that Playscii is open source, still in early development, and is offered as a pay-what-you-want download here on itch. Testing and bug reports are appreciated!
The Pannini projection is a mathematical rule for constructing perspective images with very wide fields of view. It is named in honor of Gian Paolo Pannini, an 18th Century Roman painter and professor of perspective, who may very well have used it to draw spectacular views such as the one above; for it can be realized with drawing instruments almost as easily as the standard rectilinear perspective projection. However it is not now taught in art schools, and was apparently never described in print before its recent rediscovery by a team of open source software developers.
Fragmentarium is an open source, cross-platform IDE for exploring pixel based graphics on the GPU. It is inspired by Adobe's Pixel Bender, but uses GLSL, and is created specifically with fractals and generative systems in mind.
We describe a novel algorithm for extracting a resolution-independent vector representation from pixel art images, which enables magnifying the results by an arbitrary amount without image degradation. Our algorithm resolves pixel-scale features in the input and converts them into regions with smoothly varying shading that are crisply separated by piecewise-smooth contour curves. In the original image, pixels are represented on a square pixel lattice, where diagonal neighbors are only connected through a single point. This causes thin features to become visually disconnected under magnification by conventional means, and it causes connectedness and separation of diagonal neighbors to be ambiguous. The key to our algorithm is in resolving these ambiguities. This enables us to reshape the pixel cells so that neighboring pixels belonging to the same feature are connected through edges, thereby preserving the feature connectivity under magnification. We reduce pixel aliasing artifacts and improve smoothness by fitting spline curves to contours in the image and optimizing their control points.
Le papier permanent
par Astrid Brandt
La première norme internationale pour le papier permanent (ISO 9706), publiée par l’International Organization for Standardization (organisation internationale de normalisation, désignée par l’abréviation ISO) en mars 1994, fixe " les prescriptions pour qu’un papier destiné à l’établissement de documents soit permanent ", c’est-à-dire qu’il reste chimiquement et physiquement stable pendant une longue période. Cette norme internationale est l’équivalent de la norme américaine ANSI Z39.48 de 1992 : " Permanence of paper for printed library materials ".
Pour qu’un papier puisse être déclaré conforme à la norme ISO 9706 (ou ANSI Z39.48), il doit répondre aux critères suivants :
- le pH de l’extrait aqueux de la pâte à papier doit être compris entre 7,5 et 10 ;
- l’indice Kappa de la pâte à papier, qui indique la résistance à l’oxydation (liée à la présence de lignine), doit être inférieur à 5 ;
- la réserve alcaline doit être supérieure ou égale à 2 % d’équivalent de carbonate de calcium ;
- la résistance à la déchirure doit être supérieure à 350 mN pour un papier dont le grammage est supérieur à 70 g/m2.
Le symbole attaché à cette norme est le signe mathématique de l’infini dans un cercle surmontant la mention " ISO 9706 ".
Le papier permanent peut être fabriqué soit à partir de chiffons, soit à partir de pâte chimique de bois en milieu neutre ou alcalin ; le bois peut donc être utilisé à condition d’en éliminer tous les constituants non cellulosiques, et en particulier la lignine.
Amongst the many freely available pieces of art released by Getty are a number of quite famous images, including work by Claude Monet, Vincent van Gogh, Rembrandt, and Leonardo da Vinci. The 4,600 pieces of artwork available are just the beginning, as well. Getty says that it's actively exploring the possibility of releasing much more art into the public domain, both from the museum's collection as well as materials from the Getty Research Institute's special collections. While Getty isn't the first museum to push forward with an open artwork initiative (the museum cited a number of institutions like the Walters Art Museum as inspirations for the movement), it's the latest example of how the internet is making classic, famous works more accessible.
Born in 1982. His works, centralising in real-time processed, computer programmed audio visual installations, have been shown at national and international art exhibitions as well as the Media Art Festivals. He is a recipient of many awards including the Excellence Prize at the Japan Media Art Festival in 2004, and the Award of Distinction at Prix Ars Electronica in 2008. Having been involved in a wide range of activities, he has worked on a concert piece production for Ryoji Ikeda, collaborated with Yoshihide Otomo, Yuki Kimura and Benedict Drew, participated in the Lexus Art Exhibition at Milan Design week. and has started live performance as Typingmonkeys.
History of Art, browse By Movement.
The last decade has seen an incredible growth in the production and distribution of images. The availability of unexpensive production means made amateur creativity increase exponentially, while the Internet provided a new platform of distribution for this kind of production, usually kept as private so far. In the meantime, videogames, virtual worlds and systems such as Google Street View provided this mass of prosumers with whole worlds that can be built, implemented through their own creative practice, documented and used as tools for the development of new images and new narratives. What is the impact of this process on art practice and on the artist – in the past, the only blessed depositary of the creative gesture? Which kind of dialogue is going on between amateur practices and codified languages?
Imagine you're a game producer in the late 1980s, a week before the deadline and you still haven't got a cover for your game. Exhausted from crunchtime, you tell your illustrator to just rip off some Schwarzenegger action movie to get the job done. Careful, your subordinate might take the order all too literally! When artwork in video games seems to look too realistic to be actually drawn by the artist, then it actually might be too realistic, as many vintage games have stolen images from movies, album covers, paintings and even other games. The subject here aren't simply inspired designs or characters (in that case, we'd be here all day just counting the games influenced by Nausicaä, Hokuto no Ken or Alien), but actual specific images that might have been traced, digitized or just used as direct reference. This first page is reserved for print material that goes with a game release (covers, flyers, manuals, etc.), while on the next page we'll be diving into the games themselves. Some of these are well known, others more obscure, but they all have something in common: They would likely have gotten their artists sued if the original images' copyright holders had ever seen them; a gallery of litigations that could have been, so to speak.
Halide is a new programming language designed to make it easier to write high-performance image processing code on modern machines. Its current front end is an embedding in C++
Image Atlas investigates cultural differences and similarities by indexing top image results for given search terms across local engines throughout the world. Visitors can refine or expand their comparisons from the 57 countries currently available, and sort by Gross Domestic Product (GDP) or alphabetical order.
Associations with color are defined, in part by Faber Birren (the author of Color Psychology and Color Therapy) by our senses, language, objects (or forms), and personality characteristics. color conveys moods which attach themselves to human feelings and our psychic make-up in an almost automatic fashion. This section presents the results of color associations and how they compare to other published studies.
Gifsicle is a command-line tool for creating, editing, and getting information about GIF images and animations. Making a GIF animation with gifsicle is easy:
gifsicle --delay=10 --loop *.gif > anim.gif
Extracting framesfrom animations is easy too:
gifsicle anim.gif '#0' > firstframe.gif
You can also editanimations by replacing, deleting, or inserting frames:
gifsicle -b anim.gif --replace '#0' new.gif
Some more gifsicle features:
- Batch mode for changing GIFs in place.
- Prints detailed information about GIFs, including comments.
- Control over interlacing, comments, looping, transparency...
- Creates well-behaved GIFs: removes redundant colors, only uses local color tables if it absolutely has to (local color tables waste space and can cause viewing artifacts), etc.
- It can shrink colormaps and change images to use the Web-safe palette (or any colormap you choose).
- Optimize your animations! This stores only the changed portion of each frame, and can radically shrink your GIFs. You can also use transparency to make them even smaller. Gifsicle’s optimizer is pretty powerful, and usually reduces animations to within a couple bytes of the best commercial optimizers.
- Unoptimizing animations, which makes them easier to edit.
- A dumb name.
Two other programs are included with gifsicle: gifview is a lightweight animated-GIF viewer which can show animations as slideshows or in real time, and gifdiff compares two GIFs for identical visual appearance.
Manuel français : http://www.traduction.cc/traduction/Manuel-Gifsicle-12.html
How do Adobe Photoshop, Micrografx PicturePublisher and Pegtop XFader blend layers?
This article tries to give a survey of the most important blend modes, their advantages and disadvantages, and how they are coded.
Make your Franklin is international, bearer of a cultural reflexion.
With this mind, Make your Franklin suggest each of you to
suggest each of you to re-create a symbol of modern society : the 100$
A Wayback Machine for Social Media
Explore museums from around the world, discover and view hundreds of artworks at incredible zoom levels, and even create and share your own collection of masterpieces.
Dribbble is show and tell for creatives.
The full face mask Pixelhead acts as media camouflage, completely shielding the head to ensure that your face is not recognizable on photographs taken in public places without securing permission.
Large photos of pure chemical elements from the periodic table, published under a Creative Commons license.
Kim finds inspiration from the mundane and common objects around us. Her wicked knife skills and tenacious attention to detail have create a body of work that is as impressive as it is curious.
What would the Earth look like if every elevation was inverted? If oceans were mountains and vice versa? Bathymetric (ocean depth) information from NASA is recolored here using topographic information.
The Noun Project collects, organizes and adds to the highly recognizable symbols that form the world's visual language, so we may share them in a fun and meaningful way.
Ever pause and hold the handle, wondering what you might find waiting for you on the other side of the door? Like a little real-life snow globe, interior scenes are captured in real time and reflected as a miniature world for you to view before you decide
SurveillanceSaver is a screensaver for OS X and Windows that shows live images of over 400 network surveillance cameras worldwide. A haunting live soap opera.
Une synthèse simple pour connaître ce que vous êtes en droit de copier et diffuser gratuitement, sans vous plonger dans le Code de la Propriété Intellectuelle (CPI - attention, droit français uniquement). Avec une incursion facultative dans le droit
The Hand Drawn Map Association (HDMA) is an ongoing archive of user submitted maps and other interesting diagrams created by hand.
Hideous monsters devouring ships? Cryptic symbols, correctly showing storm fronts & dangerous currents
Webcams from somewhere in the world..
Post a secret on a postcard
Ubuntu software for designers and developers
Data Visualization is a transversal discipline which harnesses the immense power of visual communication in order to explain, in an understandable manner, the relationships of meaning, cause and dependency which can be found among the great abstract masse
Tracking photos tweeted in real-time (twitpic, yfrog, twitgoo, tweetphoto, mobypicture, or img.ly)
100,000+ old advertisements to explore
Artzilla.org is dedicated to the development of experimental browser software