Name
Abstract | ||
|---|---|---|
Deep Neural Networks (DNN) received a great interest from researchers due to their capability to construct robust abstract representations of heterogeneous documents in a latent subspace. Nonetheless, mere real-valued deep neural networks require an appropriate adaptation, such as the convolution process, to capture latent relations between input features. Moreover, real-valued deep neural networks reveal little in way of document internal dependencies, by only considering words or topics contained in the document as an isolate basic element. Quaternion-valued multi-layer perceptrons (QMLP), and autoencoders (QAE) have been introduced to capture such latent dependencies, alongside to represent multidimensional data. Nonetheless, a three-layered neural network does not benefit from the high abstraction capability of DNNs. The paper proposes first to extend the hyper-complex algebra to deep neural networks (QDNN) and, then, introduces pre-trained deep quaternion neural networks (QDNN-AE) with dedicated quaternion encoder-decoders (QAE). The experiments conduced on a theme identification task of spoken dialogues from the DECODA data set show, inter alia, that the QDNN-AE reaches a promising gain of 2.2% compared to the standard real-valued DNN-AE. |
Year | DOI | Venue |
|---|---|---|
2017 | 10.1109/ASRU.2017.8268978 | 2017 IEEE Automatic Speech Recognition and Understanding Workshop (ASRU) |
Keywords | Field | DocType |
Quaternions,deep neural networks,spoken language understanding,autoencoders,machine learning | Abstraction,Subspace topology,Convolution,Computer science,Quaternion,Artificial intelligence,Artificial neural network,Deep neural networks,Spoken language | Conference |
ISBN | Citations | PageRank |
978-1-5090-4789-5 | 0 | 0.34 |
References | Authors | |
0 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Titouan Parcollet | 1 | 16 | 9.23 |
| Mohamed Morchid | 2 | 84 | 22.79 |
| georges linar es | 3 | 136 | 29.55 |