Microsoft Technology Licensing, LLC

Aleksandar Tomic

Zhengyou Zhang

Chad Balling McBride

Amol Ashok Ambardekar

Kent D Cedola

Larry Marvin Wall

Boris Bobrov

George Petre

Seiichi Kato

Lu Yuan

Ivan Stojiljkovic

Jiangyu Liu

Yasufumi Shiraishi

Daniel J Moffitt

Kenji Nikaido

Daniel Eduardo Martinez

Milos Lazarevic

Dragan Slaveski

Elizabeth Jeanne Sheldon

Marija Antic

The performance of a neural network (NN) and/or deep neural network (DNN) can limited by the number of operations being performed as well as memory data management of a NN/DNN. Using vector quantization of neuron weight values, the processing of data by neurons can be optimize the number of operations as well as memory utilization to enhance the overall performance of a NN/DNN. Operatively, one or more contiguous segments of weight values can be converted into one or more vectors of arbitrary length and each of the one or more vectors can be assigned an index. The generated indexes can be stored in an exemplary vector quantization lookup table and retrieved by exemplary fast weight lookup hardware at run time on the fly as part of an exemplary data processing function of the NN as part of an inline de-quantization operation to obtain needed one or more neuron weight values.

Flexible hardware for high throughput vector dequantization with dynamic vector length and codebook size

The performance of a neural network (NN) and/or deep neural network (DNN) can limited by the number of operations being performed as well as memory data management of a NN/DNN. Using vector quantization of neuron weight values, the processing of data by neurons can be optimize the number of operations as well as memory utilization to enhance the overall performance of a NN/DNN. Operatively, one or more contiguous segments of weight values can be converted into one or more vectors of arbitrary length and each of the one or more vectors can be assigned an index. The generated indexes can be stored in an exemplary vector quantization lookup table and retrieved by exemplary fast weight lookup hardware at run time on the flyas part of an exemplary data processing function of the NN as part of an inline de-quantization operation to obtain needed one or more neuron weight values.

A color coding engine and a comparison engine are provided. A color coding engine may be utilized to detect logical layout object attributes in a flow format document and apply a unique color to textual elements associated with each logical layout object attribute. The resulting color coded document may be saved as a target flow format document. The target flow format document may be converted to a fixed format document and then converted by a conversion engine to a flow format document. The resulting converted flow format document may be saved as an output flow format document. A comparison engine may be utilized to compare the output flow format document and the target flow format document to determine if layout information has been properly preserved in the document conversion process.

Color coding of layout structure elements in a flow format document

Image processing for productivity applications is provided. An image may be received by a computing device. The computing device may detect the edges comprising the received image and adjust the image based on a skew state of the detected edges. The computing device may then process the adjusted image to correct imbalances. The computing device may then assign an image classification to the processed image. The computing device may then adjust the processed image based on the assigned image classification.

Image processing for productivity applications

Growing community of inventors

Belgrade, Republic of Serbia

Aleksandar Tomic