JSPS Science Dialog Program Kofu Higashi High School

Size: px

Start display at page:

Download "JSPS Science Dialog Program Kofu Higashi High School"

Gavin McGee
5 years ago
Views:

1 JSPS Science Dialog Program Kofu Higashi High School July 27th, 2011 Quang-Cuong Pham JSPS postdoctoral fellow Nakamura-Takano Laboratory Department of Mechano-Informatics University of Tokyo With the assistance of Mr Satoshi Otsuki (University of Tokyo)

2 Outline of the talk 1)My itinerary: from Vietnam to France, to Japan Itinerary: 経歴 2)Research in motion planning Motion planning: 動作計画 a) Some recent applications of robotics b) What is motion planning? c) A little experiment 3)Some cultural aspects a) Vietnam b) France c) My images of Japan Research: 研究 Application: 応用 Experiment : 実験 Culture: 文化

3 1) My itinerary

4 Timeline Elementary school: 小学校 Junior high school: 中学校 High school: 高校 Master's degree: Doctoral degree (PhD): Postdoctoral researcher: 修士博士課程ポストドクター研究員

5 Studies - At high school: I liked studying science in general, especially mathematics Science: 科学 - University: I graduated from École Normale Supérieure (in Paris), an university that trains professors and researchers in science and literature Literature: 文学 - Bachelor degree: I studied mathematics, physics, computer science - Master's degree in computer science (mostly algorithms) - PhD: I studied human movements using mathematical methods Mathematics: 数学 Physics: 物理 Computer science: 情報科学 Algorithms: アルゴリズム Human movements: 人体動作

6 Research experiences While at University, I had several research experiences 1) At 3rd year, I spent the summer in a lab studying telecommunication networks 2) At 4th year, I spent a semester in an university in the USA studying biological networks 3) In my Master year, I spent a semester in a lab studying human movements (then I stayed in this lab for PhD) What I like in research is that I can create new knowledge (and also travel around the world) Laboratory: 研究室 Telecomunication: 遠距離通信 Network: ネットワーク Knowledge: 知識 Travel around the world: 世界中を回る

My current research 1)Modeling the human body To model: モデル化する Human body: 人体 Biology: 生物学 Diagnosis: 診断 Malfunction: 機能不全 We use physics, biology, and computer algorithms to model the

7 My current research 1)Modeling the human body To model: モデル化する Human body: 人体 Biology: 生物学 Diagnosis: 診断 Malfunction: 機能不全 We use physics, biology, and computer algorithms to model the human body. This can be used to better understand the human body, to diagnose physical malfunction or for sport education 2)Motion planning (topic of my talk today) Motion planning: 動作計画

8 2) Research in motion planning

9 Some recent applications Some videos of recent applications of robotics 1)Humanoid robots 2)Robotic surgery 3)Self-driving car Applications: 応用 Humanoid: ヒト型 Surgery: 外科手術 Self-driving car: 自動運転

10 The research behind these applications Example: some of the research topics behind the self-driving car: 1)Image analysis: are there any obstacles? what are the obstacles? (pedestrians, other cars, wall,...) 2)Localization: where am I? where are the obstacles? 3)Motion planning: how to reach the goal without hitting obstacles? Image analysis: 画像解析 Pedestrians: 通行人 Obstacles: 障害物 Localization: 自己位置同定

11 Image analysis The car must be able to detect automatically, from its cameras and other sensors (laser) 1)pedestrians Image analysis: 画像解析 Detect: 検出する Automatic: 自動 Pedestrians: 通行人 Road borders: 車線 2)other cars 3)road borders 4)road signs, traffic lights Road signs: 道路標識 Trafic lights: 信号

12 Localization The car must be able to localize itself with respect to the obstacles (pedestrians, other cars, road borders,...) and in space, using for example 1)Image analysis (3D vision) Localization: 自己位置同定 Integration: 積分 Speed: 速さ 2)Integration of speed, acceleration data Acceleration: 加速度 3)GPS (Global Positioning System) Combination: 組み合わせ 4)Combination of all these data

13 Motion planning The piano mover problem: how to move the piano from the starting position to the goal without touching the obstacles? Starting position: スタート地点 Goal: ゴール Obstacles: 障害物

14 Algorithm to solve this problem Step 1: select some points randomly in the free space Select: 選択する Randomly: 無作為に Free space: 自由空間

15 Algorithm (2) Step 2: join two points if one can go from one point to the other without touching the obstacles This creates a graph Join: つなぐ Graph: グラフ

16 Algorithm (3) Step 3: find a path in the graph Path: 軌道

17 More difficult: the car Why the green path is not possible for the car? Because the car cannot move sideways This is known as a nonholonomic constraint

18 The car (2) The blue path is possible for the car : it respects the nonholonomic constraints But finding the blue path is much more difficult than finding the green path!

the needle: the needle makes a circle in space 2) Turn the needle

19 Yet more difficult: the bevel needle The bevel needle is used in robotic surgery We can use two actions to drive the needle: 1) Push the needle: the needle makes a circle in space 2) Turn the needle around its axis Bevel needle: 注射針 Robotic surgery: ロボット手術 Push: 押す Circle: 円 Turn: 回す

20 Experiment How to drive the bevel needle from one position (the insertion point) to another position (where surgery is needed) while avoiding obstacles (arteries, organs,...)? The bevel needle is also a nonholonomic system Let's do an experiment using MATLAB Insertion: 挿入 Artery: 動脈 Organ: 臓器

21 Conclusion Behind the impressive applications we have seen at the beginning, there are many interesting and challenging research problems. To solve these problems, one needs mathematics and algorithmics. Impressive: 素晴らしい Challenging: 挑戦的 Mathematics: 数学 Algorithmics: アルゴリズム

22 3) Some cultural aspects

Vietnam (famous cities) Vietnam: 87 millions of inhabitants, 330.

millions people) 2) Halong: beautiful bay 3) Huê: ancient capital city 4)

23 Vietnam (famous cities) Vietnam: 87 millions of inhabitants, km2 Language: vietnamese 1) Hanoi: capital city, and my hometown (6 millions people) 2) Halong: beautiful bay 3) Huê: ancient capital city 4) Ho Chi Minh city: most populous city in Vietnam (7 millions people) Capital city: 首都 Bay: 湾

carrot, rolled in rice paper and then fried) 3) Canh chua: light sour soup made

24 Vietnam (some typical dishes) 1) Phở ( フォー ): rice noodle soup (often with beef). From the North of Vietnam 2) Nem: fried springroll (with ground meat, mushroom, carrot, rolled in rice paper and then fried) 3) Canh chua: light sour soup made with fish, pineapple, tamarind,... From the South Soup: スープ Fried: フライド Sour: 酸っぱい

25 France 1) Eating well is very important for the French people. Usually, people spend >1h at lunch, then have coffee, discuss about social and political questions,... before going back to work 2) Typical specialities: wine, cheese, sausage, foie gras, etc. Each region has its own speciality. 3) Legal working duration in France is 35 hours/week, and usually people don't do overwork. Also, people take all the vacations they have (5 weeks/year + many national holidays) 4) The French people are also very proud of their culture: literature (the most Nobel prizes in literature; each year in November, there are more than 700 new novels), museums (Le Louvre, Musée d'orsay,...), architecture, etc. Overwork: 残業

26 My images of Japan When I was at elementary school in Vietnam

27 My images of Japan (2) Then, at junior high school and high school in France

28 My images of Japan (3) Later, at University

29 My images of Japan (4) Now I have the opportunity to taste authentic Japanese food, drink sake, learn about Japanese culture, visit the places I have seen in films or read of in novels,... and meet and discuss with Japanese people! Authentic: 本物の

30 Thank you for your attention! Let's now discuss more interactively Don't hesitate to ask questions You can reach me by My website: Interactive: 対話

車載カメラにおける信号機認識および危険運転イベント検知 Traffic Light Recognition and Detection of Dangerous Driving Events from Surveillance Video of Vehicle Camera

車載カメラにおける信号機認識および危険運転イベント検知 Traffic Light Recognition and Detection of Dangerous Driving Events from Surveillance Video of Vehicle Camera * 関海克 * 笠原亮介 * 矢野友章 Haike GUAN Ryosuke KASAHARA Tomoaki YANO 要旨