| Su | Mo | Tu | We | Th | Fr | Sa |
|---|---|---|---|---|---|---|
| 1 | 2 | 3 | ||||
| 4 | 5 | 6 | 7 | 8 | 9 | 10 |
| 11 | 12 | 13 | 14 | 15 | 16 | 17 |
| 18 | 19 | 20 | 21 | 22 | 23 | 24 |
| 25 | 26 | 27 | 28 | 29 | 30 | 31 |

カーネギーメロン大学のプレスリリースより。これは面白い。左上の写真を喰わせると、そこから3Dオブジェクトを再現して、右上や右下や左下のような画像を生成できるらしい。精度にもよるけど、応用範囲広そうだなあ。
Using machine learning techniques, Robotics Institute researchers Alexei Efros and Martial Hebert, along with graduate student Derek Hoiem, have taught computers how to spot the visual cues that differentiate between vertical surfaces and horizontal surfaces in photographs of outdoor scenes. They've even developed a program that allows the computer to automatically generate 3-D reconstructions of scenes based on a single image.
ロボット工学の研究者であるAlexei EfrosとMartial Hebertは、大学院生のDerek Hoiemとともに、屋外の風景を撮った写真の垂直な面と水平な面を区別する……virtual cues? なんのことだ。 ごめんテクニカルタームが多くて訳せない。人工知能を使ってるということだけは分かったけど:-p つーか金出武雄先生がコメントしてるじゃん。わお。
どうも実世界のコンテキストを学習させておくことにより制約条件を与えることができるのがポイントらしい。事前に風景写真を大量に喰わせて「空は青い」とか「建物は地面に建っている(空に浮いていることはない)」とかいうルールを学習させておくのが強いらしい……のだがいまいちよーわからんな。まあいいや。よく知りたい人は原文をどうぞ。あと今月開催されるIEEEほげほげで発表されるとのことなんで、カンファレンスペーパーをゲットするととても良く分かると思われます。
Carnegie Mellon Researchers Teach Computers To Perceive Three Dimensions in 2-D Images
http://www.cmu.edu/PR/releases06/060613_3d.html

