Accelerating deep neural networks (DNNs) has been attracting increasing attention as it can benefit a wide range of applications, e.g., enabling mobile systems with limited computing resources to own powerful visual recognition ability. A practical strategy to this goal usually relies on a two-stage process: operating on the trained DNNs (e.g., approximating the convolutional filters with tenso...

today, with the advanced statistical techniques and neural networks, predictive models of distribution have been rapidly developed in ecology. purpose of this research is to predict and map the distribution of tetranychus urticae koch (acari: tetranychidae) using mlp neural networks combined with genetic algorithm in surface of farm. population data of pest was obtained in 2016 by sampling in 1...

This paper introduces THUMT, an opensource toolkit for neural machine translation (NMT) developed by the Natural Language Processing Group at Tsinghua University. THUMT implements the standard attention-based encoder-decoder framework on top of Theano and supports three training criteria: maximum likelihood estimation, minimum risk training, and semi-supervised training. It features a visualiza...

Supervised multi-channel audio source separation requires extracting useful spectral, temporal, and spatial features from the mixed signals. The success of many existing systems is therefore largely dependent on the choice of features used for training. In this work, we introduce a novel multi-channel, multiresolution convolutional auto-encoder neural network that works on raw time-domain signa...

Feature Extraction Features are extracted using a two-layer stacked auto-encoder (SAE) with 1440 input neurons, 500 first hidden layer neurons and 100 2nd layer hidden neurons. Each CQT input patch consists 8 frames with 180 frequency bins or one measure of the song and each chroma feature input consists of 144 frequency bins and 10 frames with 50% overlap. The SAE is then trained on a set of c...

conclusions as we can see the ann outputs values are very close to actual cu concentration, so indicating that predicted values are accurate and the network design is proper and the input variables well suitable for the prediction of cu concentration. background access to safe drinking water is one of the basic human rights and essential for healthy life. concerns about the effects of copper on...

In this paper, we propose an efficient algorithm to directly restore a clear image from a hazy input. The proposed algorithm hinges on an end-to-end trainable neural network that consists of an encoder and a decoder. The encoder is exploited to capture the context of the derived input images, while the decoder is employed to estimate the contribution of each input to the final dehazed result us...

Source separation and other audio applications have traditionally relied on the use of short-time Fourier transforms as a front-end frequency domain representation step. We present an auto-encoder neural network that can act as an equivalent to short-time front-end transforms. We demonstrate the ability of the network to learn optimal, real-valued basis functions directly from the raw waveform ...