Inspirational Machine Learning Papers from Other Fields

(a very brief selection of the flood of research in vision/voice/etc)

Factored Spatial and Spectral Multichannel Raw Waveform CLDNNS

Learning the Speech Front-end With Raw Waveform CLDNNs

Both of these works on CLDNNs are an excellent example of how learning from raw time series waveforms, building recurrent time series representations, and then mapping to supervised targets can be done with a CLDNN architecture.

Spatial Transformer Networks

A powerful method for end to end training of localization networks in the image domain by fitting 2D Affine transform parameters through a regression task prior to the discriminative task.

SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient

Generative models for acoustic time-series

WaveNet: A Generative Model for Raw Audio

Google’s WaveNet model for time series acoustic voice synthesis

Image Compression

Some impressive results on learned image compression schemes.

Radio Machine Learning Papers 

An Introduction to Machine Learning Communications Systems

An in depth overview of communications system and sensing applications of ML to comms.

Communication System PHY Learning

Learning to Communicate: Channel Auto-encoders, Domain Specific Regularizers, and Attention

The fundamentals of posing an autoencoder as a communications system.

On Deep Learning-Based Channel Decoding

Work looking into the error correction decoding capacity of deep neural networks.

Scaling Deep Learning-based Decoding of Polar Codes via Partitioning

Efforts to scale learned decoding techniques for Polar Codes to larger block sizes!

RNN Decoding of Linear Block Codes

RNN Decoding comparisons to BP decoders along with interesting complexity comparisons.

Radio Sensing

Convolutional Radio Modulation Recognition Networks

Modulation Recognition from a supervised learning standpoint.

Wireless Interference Identification with Convolutional Neural Networks

Using a similar CNN classifier technique to identify 802.11 PHY variants over the air quite successfully.

Semi-Supervised Radio Signal Identification

Scaling RF labeling beyond supervised datasets.

Rigorous Moment-Based Automatic Modulation Classification

Using DNNs on top of well regarded moment based features for modulation classification.

Radio Structure Learning

Unsupervised Representation Learning of Structured Radio Communications Signals

Unsupervised structure learning from radio datasets

Radio Transformer Networks: Attention Models for Learning to Synchronize in Wireless Systems

Augmenting learning models with just-enough expert knowledge of the physical world.