We present a method to separate speech signals from noisy environments in the embedding space of a neural audio codec. We introduce a new training procedure that allows our model to produce structured encodings of audio waveforms given by embedding vectors, where one part of the embedding vector represents the speech signal, and the rest represent the environment. We achieve this by partitioning the embeddings of different input waveforms and training the model to faithfully reconstruct audio from mixed partitions, thereby ensuring each partition encodes a separate audio attribute. As use cases, we demonstrate the separation of speech from background noise or from reverberation characteristics. Our method also allows for targeted adjustments of the audio output characteristics.

We presentSoundStream, a novel neural audio codec that can efficiently compress speech, music and general audio at bitrates normally targeted by speech-tailored codecs.SoundStreamrelies on a model architecture composed by a fully convolutional encoder/decoder network and a residual vector quantizer, which are trained jointly end-to-end. Training leverages recent advances in text-to-speech and speech enhancement, which combine adversarial and reconstruction losses to allow the generation of high-quality audio content from quantized embeddings. By training with structured dropout applied to quantizer layers, a single model can operate across variable bitrates from 3 kbps to 18 kbps, with a negligible quality loss when compared with models trained at fixed bitrates. In addition, the model is amenable to a low latency implementation, which supports streamable inference and runs in real time on a smartphone CPU. In subjective evaluations using audio at 24 kHz sampling rate,SoundStreamat 3 kbps outperforms Opus at 12 kbps and approaches EVS at 9.6 kbps. Moreover, we are able to perform joint compression and enhancement either at the encoder or at the decoder side with no additional latency, which we demonstrate through background noise suppression for speech.

Research on the Egyptian food security, has become the subject of countless studies and debates. The gap between the Egyptian domestic food production and consumption is translated into high import costs. In this paper, modeling, simulation analytical capability, expert experiences and imagination, and policy/decision makers' insights are integrated in a decision support system (DSS). The developed DSS is used to anticipate the future of the national food security and to create a consensus about the required policies that aid to achieve policy/decision makers' long-term strategic objectives. Moreover, this research is a research point in the research plan of the Egyptian agriculture research center. The gap between domestic milk production and consumption is discussed as a case-study.

Research on the future of the Egyptian tourism industry has become the subject of countless studies and debates. Different wildcards have high impacts on this crucial industry and ultimately of the future of the Egyptian national income. This paper is about an Intelligent Decision Support System (DSS) that specifically focuses on assessing the impacts of wildcards on the future revenues of the Egyptian tourism industry. The developed DSS integrates both qualitative and quantitative DSS methods (called the enhanced Trend Impact Analysis and Real-Time Delphi Survey). Also, the developed analytical tool is used to generate various scenarios (not just a single forecast) by modifying the extrapolation of historical trends in view of tourism experts.