hamadakoichiのブックマーク / 2015年12月12日

hamadakoichi id:hamadakoichi

2015年12月12日のブックマーク (7件)

Generating Images from Captions with Attention
Motivated by the recent progress in generative models, we introduce a model that generates images from natural language descriptions. The proposed model iteratively draws patches on a canvas, while attending to the relevant words in the description. After training on Microsoft COCO, we compare our model with several baseline generative models on image generation and retrieval tasks. We demonstrate
hamadakoichi 2015/12/12
リンク
Unitary Evolution Recurrent Neural Networks
- 2 users
- arxiv.org
- 学び
Recurrent neural networks (RNNs) are notoriously difficult to train. When the eigenvalues of the hidden to hidden weight matrix deviate from absolute value 1, optimization becomes difficult due to the well studied issue of vanishing and exploding gradients, especially when trying to learn long-term dependencies. To circumvent this probl em, we propose a new architecture that learns a unitary weight
hamadakoichi 2015/12/12
リンク
On Learning to Think: Algorithmic Information Theory for Novel Combinations of Reinforcement Learning Controllers and Recurrent Neural World Models
- 5 users
- arxiv.org
- 学び
This paper addresses the general probl em of reinforcement learning (RL) in partially observable environments. In 2013, our large RL recurrent neural networks (RNNs) learned from scratch to drive simulated cars from high-dimensional video input. However, real brains are more powerful in many ways. In particular, they learn a predictive model of their initially unknown environment, and somehow use i
hamadakoichi 2015/12/12
リンク
https://pageperso.lis-lab.fr/~benoit.favre/papers/favre_nips2015.pdf
hamadakoichi 2015/12/12
NIPS2015 NLP (SLU) Best Paper : "Lexical embedding adaptation for open-domain spoken language understanding"
リンク
Improved Deep Learning Baselines for Ubuntu Corpus Dialogs
hamadakoichi 2015/12/12
NIPS2015 NLP (SLU) Best Paper : "Improved Deep Learning Baselines for Ubuntu Corpus Dialogs"
リンク
Detecting Interrogative Utterances with Recurrent Neural Networks
- 1 user
- arxiv.org
- 学び
hamadakoichi 2015/12/12
NIPS2015 NLP (SLU) Best Paper : "Detecting Interrogative Utterances with Recurrent Neural Networks"
リンク
A Roadmap towards Machine Intelligence
- 9 users
- arxiv.org
- 学び
The development of intelligent machines is one of the biggest unsolved challenges in computer science. In this paper, we propose some fundamental properties these machines should have, focusing in particular on communication and learning. We discuss a simple environment that could be used to incrementally teach a machine the basics of natural-language-based communication, as a prerequisite to more
hamadakoichi 2015/12/12
リンク
- 2015年12月13日
- 2015年12月12日
- 2015年12月11日