달력

3

« 2024/3 »

  • 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

로봇고 신주희선생님 초청으로 마을 주민 아카데미 강연을 하였다. 다가오는 인공지능세상에서 우리 아이들을 어떻게 교육을 시키기고 미래를 대비해야하는지 나의 의견을 피력하였다. 전세계는 인공지능을 기반으로 경제지도가 바뀌고 있다. 구글, 아마존 등 거대 글로벌기업들은 모두 인공지능과 로봇에 막대한 자금을 투자하고, 인력자원 확보에 전력을 기울이고 있다. 우리의 혁신기업들이 이러한 인공지능기술을 개발하고, 인력을 보강하지 않으면 무한 경쟁시대에 살아남을 수 없다. 이제 인공지능을 개발단계를 넘어 활용하는 시대에 접어 들고 있다. 20년전 PC와 인터넷보급될 때처럼, 10년 스마트폰이 우리의 생활패턴을 바꾼것처럼, 10년후 인공지능기술을 활용한 제품들이 곳곳에서 새로운 변화들 주도할 것이다. 논문과 프로그램은 아카이브(Arxiv)와 깃허브(github)를 통해 무료로 공개되고 있다. 이제 마음만  먹으면 IT기술을 스스로 독학하는 시대가 열리고 있는 것이다. 인공지능기술은 나날이 발전하여 각종 교육자료와 강연자료가 유튜브를 통해 공개되고 있다. 이제 우리가 갈길은 이러한 신기술에 적극 대비하는 전략을 세우고, 새로운 패러다임의 변화 우리자신을 적극 대응하는 것이다. 적극 공감의 박수를 받으며 마무리~

 

영상보러가기

https://www.youtube.com/watch?v=luRMKBCmaZs&t=1069s

 

 

:
Posted by 고경철(kckoh)










[181115] 2018 스마트 팩토리 HR Insight 세미나


본 강의는 여의도 중소기업중앙회관에서 2018.11.15(목) 15:00 에 있었던 강연으로 4차산업혁명시대 중소기업 인력양성 방안에 대한 화두를 던지고 있다. 인공지능 전문교육기관을 통해 누구나 AI를 공부하고 활용할 수 있는 환경을 구축해야 한다.






:
Posted by 고경철(kckoh)
2018. 11. 13. 10:40

Learning to grasp - Jacob Varley 인공지능로봇/인공지능2018. 11. 13. 10:40

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Sciences - Columbia University, 2018


Abstract
Providing robots with the ability to grasp objects has, despite of research, remained a challenging problem. The problem is approachable in constrained environments where there is ample prior knowledge of the scene and objects that will be manipulated. The challenge is in building systems that scale beyond specific situational instances and gracefully operate in novel condition. In the past, heuristic and simple rule based strategies were used to accomplish tasks such as scene segmentation or reasoning about  occlusion. These heuristic strategies work in constrained environments where a roboticist can make simplifying assumptions about everything from the geometries of the objects to be interacted with, level of clutter, camera position, lighting, and a myriad of other relevant variables. With these assumptions in place, it becomes tractable for a roboticist to hardcode desired behavior and build a robotic system capable of completing repetitive tasks. These hardcoded behavior will quickly fail if the assumptions about the environment are invalidated. In this thesis, we will demonstrate how a robust grasping system can be built that is capable of operating under a more variable set of conditions without requiring significant engineering of behavior by a roboticist. 



 This robustness is enabled by a new found ability to empower novel machine learning techniques with massive amounts of synthetic training data. The ability of simulators to create realistic sensory data enables the generation of massive corpora of labeled training data for various grasping related tasks. The use of simulation allows for the creation of a wide variety of environments and experiences exposing the robotic system to a large number of scenarios before ever operating in the real world. This thesis demonstrates that it is now possible to build systems that work in the real world trained using deep learning on synthetic data. The sheer volume of data that can be produced via simulation enables the use of powerful deep learning techniques whose performance scales with the amount of data available. This thesis will explore how deep learning and other techniques can be used to encode these massive datasets for efficient runtime use. The ability to train and test of synthetic data allows for quick iterative development of new perception, planning and grasp execution algorithms that work in a large number of environments. Creative applications of machine learning and massive synthetic datasets are allowing robotic systems to learn skills, and move beyond repetitive hardcoded tasks. 



일단 이논문은 파지작업과 CNN을 결합한 것으로 보인다. RL기반은 아닌 것이다. 특이한 것은 에너지 최적화를 주제삼은 것이다. 잡는냐 못잡는냐보다 접촉에너지를 최소화하는게 왜 중요한 문제인지 이해가 안간다. Heatmap이 신경망의 최종출력인데, 손바닥의 터치 등고선을 고려한 것일까. 키워드는 포인트 클라우드, RGBD 키넥트센서, 

일단 우리의 작업을 위해 일단 물체를 환경으로 부터 분리하는 Semantic map을 구성하는 능력이 필요하다. 이는 RGB영상처리에서 RGBD영상처리에 대한 연구가 필요한 것이다. 

:
Posted by 고경철(kckoh)