We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
One-shotで模倣学習を、しかもドメインを変えて行う研究。具体的には(ロボットのでなく)人の動作を一回見て模倣する。事前学習として人とロボットの動作のペア画像を用意し、戦略の初期値を人の画像から、初期値からの学習をロボットの画像から行う。これで人の画像から良い初期値を得る関数を学習する
https://arxiv.org/abs/1802.01557
Tianhe Yu, Chelsea Finn, Annie Xie, Sudeep Dasari, Tianhao Zhang, Pieter Abbeel, Sergey Levine
2018/2/5
The text was updated successfully, but these errors were encountered:
この前段階の研究として、ドメイン転化無しのOne shotがある。
One-Shot Visual Imitation Learning via Meta-Learning (こちらについては実装が公開されている: https://github.com/tianheyu927/mil)
双方の研究が、公式ブログで紹介されている。
One-Shot Imitation from Watching Videos
Sorry, something went wrong.
解説スライド
https://www.slideshare.net/DeepLearningJP2016/oneshot-imitation-from-observing-humans-via-domainadaptive-metalearning
No branches or pull requests
一言でいうと
One-shotで模倣学習を、しかもドメインを変えて行う研究。具体的には(ロボットのでなく)人の動作を一回見て模倣する。事前学習として人とロボットの動作のペア画像を用意し、戦略の初期値を人の画像から、初期値からの学習をロボットの画像から行う。これで人の画像から良い初期値を得る関数を学習する
論文リンク
https://arxiv.org/abs/1802.01557
著者/所属機関
Tianhe Yu, Chelsea Finn, Annie Xie, Sudeep Dasari, Tianhao Zhang, Pieter Abbeel, Sergey Levine
投稿日付(yyyy/MM/dd)
2018/2/5
概要
新規性・差分
手法
結果
コメント
The text was updated successfully, but these errors were encountered: