Non-linear manifold learning enables high-dimensional data analysis, but requires outof-sample-extension methods to process new data points. In this paper, we propose a manifold learning algorithm based on deep learning to create an encoder, which maps a high-dimensional dataset and its low-dimensional embedding, and a decoder, which takes the embedded data back to the high-dimensional space. S...

The IS-95 Direct Sequence Code Division Multiple Access (DS-CDMA) system has become a U.S. digital cellular standard. In the forward traffic channels (from the base station to the mobiles), the speech encoder rate can vary, depending on the speech activity in the transmitter. At the receiver, the decoder does not know which one of the speech encoder rate the transmitter used. This paper focuses...

We propose a novel extension of the encoder-decoder framework, called a review network. The review network is generic and can enhance any existing encoderdecoder model: in this paper, we consider RNN decoders with both CNN and RNN encoders. The review network performs a number of review steps with attention mechanism on the encoder hidden states, and outputs a thought vector after each review s...

In this paper, the performance of three types of serial concatenated convolutional codes (SCCC) is compared and analyzed in additive white Gaussian noise (AWGN) channel. In Type I, only the parity bits of outer encoder are passed to inner encoder. In Type II and Type III, both the information bits and the parity bits of outer encoder are transferred to inner encoder. As results of simulation, T...

An Improved Adaptive Arithmetic Coding is presented in the paper to be applicable in contemporary hybrid video encoders. The proposed Adaptive Arithmetic Encoder makes the improvement of the state-of-the-art Context-based Adaptive Binary Arithmetic Coding (CABAC) method that is used in the new worldwide video compression standard MPEG-4 AVC/H.264. The improvements were obtained by the use of mo...

We propose a selective encoding model to extend the sequence-to-sequence framework for abstractive sentence summarization. It consists of a sentence encoder, a selective gate network, and an attention equipped decoder. The sentence encoder and decoder are built with recurrent neural networks. The selective gate network constructs a second level sentence representation by controlling the informa...

The increase popularity of multimedia application especially in image processing places a great demand on efficient data storage and transmission techniques. Network communication such as wireless network can easily be intercepted and cause of confidential information leaked. Unfortunately, conventional compression and encryption methods are too slow; it is impossible to carry out real time sec...

In this project a servo (position) controller that is able to receive commands via a Controller Area Network (CAN) is presented. Additionally, the board was developed to be as generally usable as possible in the sense that if the DC motor to be controlled does not have a rotary encoder, then a contactless magnetic rotary encoder (CMRE) module can be plugged into the board (via two SIL connector...