Abstract: A method of operating a neural network includes determining a complexity, such as a number) of separable filters approximating a filter. The method further includes selectively applying a decomposed convolution to the filter based on the determined number of separable filters.

Abstract: A method for selecting a reduced number of model neurons in a neural network includes generating a first sparse set of non-zero decoding vectors. Each of the decoding vector is associated with a synapse between a first neuron layer and a second neuron layer. The method further includes implementing the neural network only with selected model neurons in the first neuron layer associated with the non-zero decoding vectors.

Abstract: Systems and methods are provided for distributed video storage and search with edge computing. The method may comprise caching a first portion of data on a first device. The method may further comprise determining, at a second device, whether the first device has the first portion of data. The determining may be based on whether the first piece of data satisfies a specified criterion. The method may further comprise sending the data, or a portion of the data, and/or a representation of the data from the first device to a third device.

Abstract: A method for selecting bit widths for a fixed point machine learning model includes evaluating a sensitivity of model accuracy to bit widths at each computational stage of the model. The method also includes selecting a bit width for parameters, and/or intermediate calculations in the computational stages of the mode. The bit width for the parameters and the bit width for the intermediate calculations may be different. The selected bit width may be determined based on the sensitivity evaluation.

Abstract: Compressing a machine learning network, such as a neural network, includes replacing one layer in the neural network with compressed layers to produce the compressed network. The compressed network may be fine-tuned by updating weight values in the compressed layer(s).

Abstract: The present disclosure provides systems and methods for mapping a determined path of travel. The path of travel may be mapped to a camera view of a camera affixed to a vehicle. In some embodiments, the path of travel may be mapped to another view that is based on a camera, such as a bird's eye view anchored to the camera's position at a given time. These systems and methods may determine the path of travel by incorporating data from later points in time.

Abstract: Systems, devices, and methods are provided for distributed search with edge computing. Enabled systems and devices may overcome challenges associated with searching data captured by one or more connected devices, including privacy, security, bandwidth, backhaul, and memory storage.

Abstract: A method of reducing image resolution in a deep convolutional network (DCN) includes dynamically selecting a reduction factor to be applied to an input image. The reduction factor can be selected at each layer of the DCN. The method also includes adjusting the DCN based on the reduction factor selected for each layer.

Abstract: Multi-label classification is improved by determining thresholds and/or scale factors. Selecting thresholds for multi-label classification includes sorting a set of label scores associated with a first label to create an ordered list. Precision and recall values are calculated corresponding to a set of candidate thresholds from score values. The threshold is selected from the candidate thresholds for the first label based on target precision values or recall values. A scale factor is also selected for an activation function for multi-label classification where a metric of scores within a range is calculated. The scale factor is adjusted when the metric of scores are not within the range.

Abstract: Context-based priors are utilized in machine learning networks (e.g., neural networks) for detecting objects in images. The likely locations of objects are estimated based on context labels. A machine learning network identifies a context label of an entire image. Based on the, the network selects a set of likely regions for detecting objects of interest in the image.

Abstract: A method for selecting bit widths for a fixed point machine learning model includes evaluating a sensitivity of model accuracy to bit widths at each computational stage of the model. The method also includes selecting a bit width for parameters, and/or intermediate calculations in the computational stages of the mode. The bit width for the parameters and the bit width for the intermediate calculations may be different. The selected bit width may be determined based on the sensitivity evaluation.

Abstract: A method of adaptively selecting a configuration for a machine learning process includes determining current system resources and performance specifications of a current system. A new configuration for the machine learning process is determined based at least in part on the current system resources and the performance specifications. The method also includes dynamically selecting between a current configuration and the new configuration based at least in part on the current system resources and the performance specifications.

Abstract: A method of quantizing a floating point machine learning network to obtain a fixed point machine learning network using a quantizer may include selecting at least one moment of an input distribution of the floating point machine learning network. The method may also include determining quantizer parameters for quantizing values of the floating point machine learning network based at least in part on the at least one selected moment of the input distribution of the floating point machine learning network to obtain corresponding values of the fixed point machine learning network.

Abstract: A method of reducing image resolution in a deep convolutional network (DCN) includes dynamically selecting a reduction factor to be applied to an input image. The reduction factor can be selected at each layer of the DCN. The method also includes adjusting the DCN based on the reduction factor selected for each layer.

Abstract: Compressing a machine learning network, such as a neural network, includes replacing one layer in the neural network with compressed layers to produce the compressed network. The compressed network may be fine-tuned by updating weight values in the compressed layer(s).

Abstract: A method for classifying an object includes applying multiple confidence values to multiple objects. The method also includes determining a metric based on the multiple confidence values. The method further includes determining a classification of a first object from the multiple objects based on a knowledge-graph when the metric is above a threshold.

Abstract: A method of training a neural network includes encouraging one or more filters in the neural network to have a low rank.

Abstract: A method of operating a neural network includes determining a complexity, such as a number) of separable filters approximating a filter. The method further includes selectively applying a decomposed convolution to the filter based on the determined number of separable filters.

Abstract: A method of distributed computation includes computing a first set of results in a first computational chain with a first population of processing nodes and passing the first set of results to a second population of processing nodes. The method also includes entering a first rest state with the first population of processing nodes after passing the first set of results and computing a second set of results in the first computational chain with the second population of processing nodes based on the first set of results. The method further includes passing the second set of results to the first population of processing nodes, entering a second rest state with the second population of processing nodes after passing the second set of results and orchestrating the first computational chain.

Abstract: A method for communicating a non-binary value in a spiking neural network includes encoding, with an encoder, a non-binary value as one or more spikes of at least one pre-synaptic neuron in a temporal frame. The method also includes computing a value with a decoder matched to the encoder. The value is computed by at least one post-synaptic neuron. The value is based on at least one synaptic weight and on the encoded spikes received from the pre-synaptic neuron.