Abstract: A method and apparatus for variable rate compression with a conditional autoencoder is herein provided. According to one embodiment, a method includes training a conditional autoencoder using a Lagrange multiplier and training a neural network that includes the conditional autoencoder with mixed quantization bin sizes.

Abstract: A method and apparatus for variable rate compression with a conditional autoencoder is herein provided. According to one embodiment, a method includes training a conditional autoencoder using a Lagrange multiplier and training a neural network that includes the conditional autoencoder with mixed quantization bin sizes.

Abstract: A method and apparatus for providing a rotational invariant neural network is herein disclosed. According to one embodiment, a method includes receiving a first input of an image in a first orientation and training a kernel to be symmetric such that an output corresponding to the first input is the same as an output corresponding to a second input of the image in a second orientation.

Abstract: A method and system are herein disclosed. The method includes developing a joint latent variable model having a first variable, a second variable, and a joint latent variable representing common information between the first and second variables, generating a variational posterior of the joint latent variable model, training the variational posterior, and performing inference of the first variable from the second variable based on the variational posterior.

Abstract: A method and system for constructing a convolutional neural network (CNN) model are herein disclosed. The method includes regularizing spatial domain weights, providing quantization of the spatial domain weights, pruning small or zero weights in a spatial domain, fine-tuning a quantization codebook, compressing a quantization output from the quantization codebook, and decompressing the spatial domain weights and using either sparse spatial domain convolution and sparse Winograd convolution after pruning Winograd-domain weights.

Abstract: A system and method for removing bias from a frequency estimate. A simulation is used to predict, for various values of the signal to noise ratio, a bias in a raw frequency estimate produced by a frequency estimation algorithm. A straight line is fit to simulated frequency offset estimates as a function of true frequency offset, and the reciprocal of the slope of the line is stored, as a multiplicative bias removal term, in a lookup table, for the simulated signal to noise ratio. In operation, the raw frequency estimate is multiplied by a multiplicative bias removal term, obtained from the lookup table, to form a corrected frequency offset estimate.

Abstract: A method is provided. The method includes selecting a neural network model, wherein the neural network model includes a plurality of layers, and wherein each of the plurality of layers includes weights and activations; modifying the neural network model by inserting a plurality of quantization layers within the neural network model; associating a cost function with the modified neural network model, wherein the cost function includes a first coefficient corresponding to a first regularization term, and wherein an initial value of the first coefficient is pre-defined; and training the modified neural network model to generate quantized weights for a layer by increasing the first coefficient until all weights are quantized and the first coefficient satisfies a pre-defined threshold, further including optimizing a weight scaling factor for the quantized weights and an activation scaling factor for quantized activations, and wherein the quantized weights are quantized using the optimized weight scaling factor.