論文まとめ
この記事について
自分の研究に関連する論文のまとめ．あるいは学生とのシェア用．気まぐれに随時更新．
映像から言語
- Nguyen et al., (2017). "Translating Videos to Commands for Robotic Manipulation with Deep Recurrent Neural Networks." [arXiv]
言語とロボット動作の変換
言語から動作
- Yamada et al., (2016/07). "Dynamical Integration of Language and Behavior in a Recurrent Neural Network for Human–Robot Interaction." [Frontiers]
動作から言語
- Heinrich and Wermter., (2014). "Interactive Language Understanding with Multiple Timescale Recurrent Neural Networks," ICANN2017, [Springer]
双方向変換
- Ogata et al., (2007). "Two-way translation of compound sentences and arm motions by recurrent neural networks," IROS2007. [IEEEXplore]
- Sugita and Tani., (2005). "Learning Semantic Combinatoriality from the Interaction between Linguistic and Behavioral Processes." [SAGE]
Semantic Navigation
Vision-and-Language Navigation (VLN．画像と文章から)
- Das et al., (2017/12). "Embodied Question Answering." [arXiv]
- Anderson et al., (2017/11). "Vision-and-Language Navigation: Interpreting visually-grounded navigation instructions in real environments." [arXiv]
- Hermann et al., (2017/06). "Grounded Language Learning in a Simulated 3D World." [arXiv]
- Chaplot et al., (2017/06). "Gated-Attention Architectures for Task-Oriented Language Grounding." [arXiv]
LRFによるSemantic Navigation
- Luo and Chen., (2017). "Recursive Neural Network Based Semantic Navigation of an Autonomous Mobile Robot through Understanding Human Verbal Instructions," IROS2017, No open-access file.
翻訳，主にneural machine translation (NMT)
- Lample et al., FAIR, (2017/11). "Unsupervised Machine Translation Using Monolingual Corpora Only." [arXiv]
- Johnson et al., Google, (2016/11). "Google's Multilingual Neural Machine Translation System : Enabling Zero-Shot Translation." [arXiv]
- Luong et al., Stanford Univ., (2015/08). "Effective Approaches to Attention-based Neural Machine Translation." [arXiv]
- Bahdanau et al., w/Bengio, (2014/09)."Neural Machine Translation by Jointly Learning to Align and Translate." [arXiv]
Attention関連の技術
- See et al., Stanford Univ. and Google, (2017/04). "Get To The Point: Summarization with Pointer-Generator Networks." [arXiv]
- Gu et al., (2016/03). "Incorporating Copying Mechanism in Sequence-to-Sequence Learning." [arXiv]
- Vinyals et al., Google Brain, (2015/06). "Pointer networks." [arXiv]
表現学習（Representation Learning)
- Tran et al., Open AI, (2017/12). "Feature-Matching Auto-Encoders." [pdf]
コミュニケーションによる記号創発
- Havrylov and Titov, (2017/05). "Emergence of Language with Multi-agent Games: Learning to Communicate with Sequences of Symbols." [arXiv]
- Mordatch and Abbeel, (2017/03). "Emergence of Grounded Compositional Language in Multi-Agent Populations." [arXiv]
ニューラルネット，ディープラーニング基本テクニックやモデルあれこれ
- CapsNet初出．Sabour et al., (2017/11). "Dynamic Routing Between Capsules." [arXiv]
- Adam初出．Kingma and Ba, (2014/12). "Adam: A Method for Stochastic Optimization." [arXiv]
- VAE初出．Kingma and Welling, (2013/12). "Auto-Encoding Variational Bayes." [arXiv]
- Elman型RNN. Elman, (1990). "Finding Structure in Time." [ScienceDirect]
- BPTT初出．Rumelhart et al., (1986). "“Learning internal representations by error propagation." [pdf]
レビュー論文
- DLによる自然言語処理．Young et al., (2017/08). "Recent Trends in Deep Learning Based Natural Language Processing." [arXiv]
- DLによる音楽生成．Briot et al., (2017/09). "Deep Learning Techniques for Music Generation - A Survey." [arXiv]
- 記号創発ロボティクス．Taniguchi et al. (2015/09). "Symbol Emergence in Robotics: A Survey." [arXiv]
古典
- シンボルグラウンディング, Harnad, (1990/02). "The Symbol Grounding Problem." [ScienceDirect]
- SHRDLU, Winograd, (1972/01). "Understanding natural language." [ScienceDirect]