nedelja, 17. marec 2013

Projekt "Večna plesišča"

(English version here)

Baje je v EU po smrti mogoč samo pokop ali zažig. V ZDA je mogoče tudi utekočinjenje, v Sloveniji pa naj bi bila zadnja leta mogoča še predelava v diamant. To je vse. Nemogoče je npr. zahtevati, da bi se človek po smrti dal nagačiti, in pričakovati, da bo država temu ustregla. Tako se je rodil projekt "Večna plesišča". 

Cilj projekta je, da bi plesalci lahko še po smrti večno plesali, ne na kakšnem zaslonu, ampak zares, v prvotni obliki. To je mogoče uresničiti tako, da kandidat še zaživa zajame gibe v digitalno obliko, ti pa v odprtokodnem formatu v javni zbirki podatkov čakajo na njegovo smrt, po kateri bo kandidatovo fizično telo preparirano in nagačeno, vanj pa vsajen robotski skelet. V kontrolni mehanizem skeleta bodo preneseni zajeti podatki, in nagačeno telo bo lahko poustvarjalo gibe plesalca, kakršen je bil nekoč.

Projekt je bil izbran v drugi krog na lanskem natečaju Robots and Avatars, tedaj pod naslovom 1st Stage Preparations for a Taxidermic Afterlife Party.

Eternal dancefloors v1



Projekt je ambiciozen, zato je razdeljen na več faz. Dokončana je prva - zajem gibov, njihovo shranjevanje v odprtokodnem formatu in vizualizacija množice zajetih plesalcev na virtualnem plesišču. Vse skupaj je izvedeno je s primitivno tehnologijo, ki se ji bodo v prihodnosti smejali, vendar deluje. 

Kandidat se postavi pred 3D senzor (Microsoft Kinect), ki ga prepozna kot človeka in vklopi odštevanje do začetka zajemanja podatkov. Nato steče desetsekundno zajemanje gibov v treh dimenzijah in hkratna vizualizacija. Ob kandidatu so na zaslonu prikazani plesalci, ki so ta postopek že izvedli, tako da kandidat ne nastopa sam, temveč v družbi. Tako je tudi laže plesati.

Za boljšo predstavo je postopek prikazan v spodnjem videu. Posnet je bil 15. februarja v Mariboru na festivalu MED, ki ga je organizirala Kibla in prijazno gostila naš projekt.



Vizualizacija je celotnega postopka zajema gibov enega plesalca se začne nekje proti drugi polovici videa. Video: Matej Kristovič.

Iz projektne dokumentacije:


‘1st Stage Preparations for a Taxidermic Afterlife Party’ is a part of a planned wider ‘Taxidermic Afterlife Party’ project, which is firstly addressing the problem of the disappearing intergenerational solidarity through the creation of taxidermic dancing afterlife avatars. 

As a conceptual starting-point we take the present situation, where society’s mechanisms are less and less able to provide for it’s older - i.e., “non-functional” - members. As a response to this phenomenon we strive to establish an absurd dystopian vision of a situation that has gone out of hands, where we have got real physical avatars with no reasonable purpose, but they do not want to go away (are present after individual’s life) and on top of this also need to be up kept (because we deal with real prepared human bodies - containing a dance mechanism - that need to go dancing / clubbing, as they function on the basis of the Tamagotchi principle). 

The artists themselves are of course submitting their bodies to this artistic project as an act of social comment.

Because of the fact that in our society you have got only three options of what can be done with your body (burial, cremation or liquefaction), one of the aims of this 1st stage is the assertion of the right to get prepared after death. Individuals that are taking part in this project are also signers of this claim (although you can take part and not sign the claim and vice versa). This whole vision might be dystopian in its core, but there is also something romantic in dancing just a little bit longer ...

In the history of human kind, dance is one of the oldest forms of expression, social interaction and establishment of collective identity; it was a part of first rituals, also meant to change each individual neural activity in order to reach this state of collective identity. Vanishing of this phenomenon or its limitation to club environment in today’s society on one hand, and flourishing use of social networks on the other, makes it interesting to put this ”primitive praxis” (dance) in the context of new technologies (virtual environment).

Stages of the whole lifelong and afterlife project:
- 1st: establishment of the dance moves database with visualization and interaction platform and functionality for asserting the right to get prepared after life via a petition
- 2nd: getting in touch with competent and/or suitable institutions (e. g., cyborg foundations) resulting in actual preparation
- 3rd: taxidermic afterlife party: embodiment of recorded database by actualization in a robotic platform
- 4th (“sad-but-true” future vision): you / your avatar will probably get sold on eBay, stored in some dusty garage, your dance moves are going to be hacked to sadly entertain the owner’s drunken friends … But no one is saying that the first exemplar is not going to end up in Guggenheim. 

We have a process here where the dancing human body is substituted by a digital representation (caught with motion capture) and later on the digital representation gets substituted again by the real body (prepared body with an implanted robotic mechanism). The whole project is resulting then in reversing the process where we establish an avatar through the omission of the real body and make our own personality avatar’s content – now this at one point “abandoned” dimension (i.e., the real body) becomes the avatar ...

The 1st stage of the project includes an interactive installation, where individuals record their dance moves through the usage of motion capture, and the development of an online virtual environment. This interauthorship (individuals contributing to the database of dance moves) can be seen as an investment into individual’s future presence and also as a contribution to the future presence of others, as the project is based on the creative commons principle. The database can be understood as a prospect for your own and others’ afterlife presence, but also as a part of the responsive environment, in which individuals enroll and take an active part in it in this lifetime. People would be called to get their digital dancing avatars through announcements / appeals in mass media.

The 1st stage can be interpreted as a project in itself with following outputs: 
- (world’s largest) database of freestyle party dance moves, including moves by professional dancers and supporters of the project
- online virtual environment, i.e., visualization and interaction platform for recorded dance moves
- a base of exclusive music sets contributed by well-known artists
- a formal claim for a right to get prepared after death.


 
Še en video, ki je nastal med razvojem:



Projekt sva izvedla Pina Gabrijan in Marko Plahuta. 

Tehnologija:
- 3D senzor: MS Kinect
- programska orodja: Processing core z GlGraphics za vizualizacijo in Apache SOLR za shranjevanje zajetih gibov.
 

sreda, 6. marec 2013

Kaj paše skupaj v loncu glede na omrežno analizo receptov

Včasih v kuhinji zmanjka idej, ali pa jedi zmanjka nekaj malega, kar bi jo naredilo res posebno ... če bi vedeli, kaj. Računalnik na pomoč! 

Sestavine v receptih so v posebnem razmerju. Skupaj pripomorejo k želenemu okusu, zato ni naključje, da se nekatere vedno pojavljajo skupaj. K pašti pač sodi paradižnikova omaka s česnom in baziliko, v sladice pa navadno damo čokolado, sladila in maščobe. Indijske in tajske jedi skoraj vedno vsebujejo ingver, čili in kokosovo mleko ... kombinacije se vrstijo v nedogled, vendar se nekatere ponavljajo.

Najbolje bi bilo analizirati, katere se največkrat pojavljajo skupaj in to nekako narisati. Preglednice in tabele so passe, z grafikoni se je tudi težko hvaliti. Omrežna analiza pa se ne sliši slabo.

Kaj, če bi skonstruirali graf in vanj dali vse sestavine vseh dosegljivih receptov? Sestavine v posameznem receptu so medsebojno povezane, in ko  seštejemo povezave vseh sestavin v vseh receptih in vsoto prikažemo kot omrežje, so sestavine, ki pogosteje nastopajo skupaj, prikazane bliže. Sestavine, ki nastopajo večkrat, so večje.

Celotno omrežje je videti tako:






Sliko je mogoče povečati (2x klik ali kontrolnik spodaj desno), pomanjšati in premikati.

Obdelal sem približno 22.000 receptov z normaliziranimi sestavinami (hvala, Najdi.si) in nastalo omrežje vizualiziral z Gephijem. Povezav med sestavinami je preveč, zato niso prikazane. Najpogostejše sestavine - sol, moka, voda, maslo, jajca, ... - sem moral odstraniti, ker so popolnoma pokvarile diagram.  

Človek bi pričakoval, da bodo skupine sestavin bolj izrazite, da bodo npr. jasno vidne gruče sestavin raznih etničnih kuhinj, a žal ni tako. Večina jedi pade v dve skupini: slane ali sladke. Slane se še delijo v podskupine, sladkih pa glede na omrežno geometrijo ni mogoče razdeliti v skupine tako, da bi bilo na prvi pogled smiselno, kaj te predstavljajo.

Zanimivo je videti sestavine, ki povezujejo slane in sladke jedi, npr. limone, kvas, sladka in kisla smetana in razna vina. Bananini listi in nojeva jajca na prominentnih mestih so posledica slabega delovanja algoritma za prepoznavo sestavin s spletnih strani.

Še približane slike. Najprej sladkarije (klik za povečanje):





in še slane jedi (klik za povečanje):






Orodja: Gephi in nekaj programiranja.
Podatki: Najdi.si.

Dober tek!