Kyocera Document Solutions Inc.

Kyocera Mita Corporation

Peerless Systems Corporation

Dongpei Su

Kenneth Allen Schmidt

Sheng Li

Michael M Chang

Kendrick Esperanza Wong

Masayoshi Nakamura

Stephen L Schafer

Alfred Abkarian

Larry Lubman

Kilho Shin

Yuzuru Naganuma

Christa Neil

Edwin Philip Lockwood

Thien-Phuc Nguyen Do

Donald Han

Waqas Rasheed

Guo Li

Robert Westervelt

Lakhbir Gandhi

Mark Raley

Michele A Lipman

Raymond B Robinson

Alex Horvath

A digital image comparator includes an Faster R-CNN configured to generate a filtered set of local feature maps of an input image, a match head, and logic to preserve the local feature maps in an indexed data structure and to make the local feature maps retrievable by the match head via local feature map indexes.

Computationally efficient document stamp detector

A device such as a color printer includes a main memory, a cache memory, and a convolutional neural network configured to convert pixels from a first color space to a second color space. The convolutional neural network is organized into execution-separable layers, and loaded one or more layer at a time (depending on cache size) from the main memory to the cache memory, whereby the pixels are processed through each of the layers in the cache memory, and layers that have completed processing are evicted to make room for caching next layer(s) of the network.

Complex color space conversion using convolutional neural networks

A method includes generating multiple attention map for an image from outputs of a corresponding different one of multiple convolutional hidden layers of a neural network. A different weighted attention map is then generated from each of the attention map. The weighted attention maps are input to a first fully-connected neural network layer to generate a colorfulness metric, which may be used to augment human-perceived colorfulness of the image.

Measurement and application of image colorfulness using deep learning

Embodiments provide an automated approach for generating unbiased synthesized image-label pairs for colorization training of retro photographs. Modern grayscale images with corresponding color images are translated to images with the characteristics of retro photographs, thereby producing training data that pairs images with the characteristics of retro paragraphs with corresponding color images. This training data can then be employed to train a deep learning model to colorize retro photographs more effectively.

Retro-to-modern grayscale image translation for preprocessing and data preparation of colorization

A digital image processor includes a region proposal network configured to transform digital image inputs into region proposals and bounding box refinement logic configured to transform the region proposals by determining a first set of the region proposals exhibiting dense spacing, determining a second set of the region proposals exhibiting sparse spacing, executing a first transformation to merge at least some of the region proposals exhibiting dense spacing to generate refined region proposals, executing a second transformation to join at least some of the region proposals exhibiting sparse spacing to generate additional ones of the refined region proposals, and applying an expansion transformation to the refined region proposals.

Consolidation of bounding boxes generated from text type pattern localizer

A method of detecting a digital stamp pattern involves operating a scanning device to create a page image scanned from a document page, inputting the page image into a one-shot trained neural network, the one-shot trained neural network configured to recognize a copy-guard digital stamp pattern using one-shot learning, analyzing the page image using the one-shot trained neural network to detect the copy-guard digital stamp pattern, and on condition the copy-guard digital stamp pattern is detected, issuing an electronic alert.

One-shot neural network for stamp detection

Systems and methods upscale an input image by a final upscaling factor. The systems and methods employ a first module implementing a super resolution neural network with feature extraction layers and multiple sets of upscaling layers sharing the feature extraction layers. The multiple sets of upscaling layers upscale the input image according to different respective upscaling factors to produce respective first module outputs. The systems and methods select the first module output with the respective upscaling factor closest to the final upscaling factor. If the respective upscaling factor for the selected first module output is equal to the final upscaling factor, the systems and methods output the selected first module output. Otherwise, the systems and methods provide the selected first module output to a second module that upscales the selected first module output to produce a second module output corresponding to the input image upscaled by the final upscaling factor.

Super resolution neural network with multiple outputs with different upscaling factors

Systems and methods for processing images receive an input image. The systems and methods provide the input image to a first module to increase a resolution of the input image to produce an upscaled image. The systems and methods detect white pixels in the input image. The systems and methods generate a mask associated with the input image. The mask includes mask bits that are set to mark the white pixels in the input image. The systems and methods upscale the mask to produce an upscaled mask matching a resolution of the upscaled image. The systems and methods identify target pixels of the upscaled image that correspond to the set mask bits in the upscaled mask. The systems and methods modify the upscaled image to produce an output image by replacing target pixels of the upscaled image with a replacement pixel having greater whiteness. The systems and methods output the output image.

White background protection in SRGAN based super resolution

Methods and systems for training and utilizing an artificial neural network (ANN) are provided. In an example method, a computing device could receive an input image comprising a plurality of channels and determine a saliency map for the input image. The computing device could also establish at least one of the plurality of channels as a training channel and at least some of the plurality of channels as one or more ground truth channels. Further, the computing device could train an ANN to predict one or more output channels from the one or more training channels, where the training involves computationally updating weights of the ANN based on a loss function that comprises a difference between the one or more output channels and the one or more ground truth channels, and where the difference is computationally biased based on values from the saliency map.

Use of a saliency map to train a colorization ANN

An example system includes a processor and a non-transitory computer-readable medium having stored therein instructions that are executable to cause the system to perform various functions. The functions include obtaining an image associated with a print job, and providing the image as input to a convolutional neural network. The convolutional neural network includes a residual network, upscaling layers, and classification layers configured to detect whether the image is an artificial image having a computer-generated image gradient. The functions also include determining, based on an output of the classification layers, that the image is an artificial image having a computer-generated image gradient. Further, the functions include, based on determining that the image is an artificial image having a computer-generated image gradient, providing the image to an upscaling module of a print pipeline for upscaling rather than using an output of the upscaling layers for the upscaling.

Super-resolution convolutional neural network with gradient image detection

Methods and systems for training and utilizing an artificial neural network (ANN) are provided. In an example method, a computing device can receive an image pair, where a first image of the image pair includes a training image and a second image of the image pair includes a ground truth image. The computing device can provide instances of the first image to a plurality of image filtering modules and determine respective filtered representations of the first image using the plurality of image filtering modules. The computing device can indirectly train an adaptor ANN by applying the adaptor ANN on the respective filtered representations to produce an adapted representation; determining, using a trained colorization ANN, a colorized image from the adapted representation; and updating weights of the adaptor ANN based on a loss function that comprises a difference between the colorized image and the second image of the image pair.

Convolutional neural network filter stacker for colorization preprocessing

A memory control method generates write control data by predetermined areas by attaching processing priority data and a transfer ID, generates read control data for respective predetermined areas by attaching processing priority data and a transfer ID, stores the write control data and the read control data in an arbitration queue, executes an arbitration that determines a priority order of processing taking the write control data and the read control data that are stored in the arbitration queue as targets, and changes the processing priority data every time a predetermined time elapses. The arbitration raises a priority order taking the write control data or the read control data that includes the processing priority data as a timeout request when the processing priority data has passed a preliminarily set threshold by the change.

Read/write direction-based memory bank control for imaging

Methods and apparatus for training and utilizing an artificial neural network (ANN) are provided. A computing device can receive training documents including text. The computing device can parse the training documents to determine training data items. Each training data item can include a training label related to text within the training documents and location information indicating a location of text related to the training label. An ANN can be trained to recognize text using the training data items and training input that includes the training documents. After training the ANN, a request to predict text in application documents that differ from the training documents can be received. The application documents can include second text. A prediction of the second text can be determined by applying the trained ANN to the application documents. After determining the prediction of the second text, information related to the second text can be provided.

Optical character recognition training data generation for neural networks by parsing page description language jobs

Example systems and related methods relate to rendering using smooth shading. A computing device can receive information about an image, the information including one or more patches, where a particular patch of the one or more patches is specified using a plurality of non-linear equations. The computing device can determine one or more linear approximations to a particular non-linear equation of the plurality of non-linear equations. The computing device can update the particular patch to replace the particular non-linear equation with at least one linear approximation of the one or more linear approximations. The computing device can render at least part of the image by at least rendering the updated particular patch. The computing device can generate an output that includes the rendered at least part of the image.

Utilizing smooth shading patches in image rendering

A memory control method receives a read request data set including a management ID and requesting reading of data from a memory, transmits read data that is read corresponding to the read request, stores write data temporarily in a write buffer, stores the read data read from the memory temporarily in a read buffer, stores write control data and read control data in an arbitration queue, executes an arbitration that determines a priority order of processing taking the write control data and the read control data that are stored in the arbitration queue as targets, and matches an order of transmitting a plurality of pieces of read data that are read from the memory corresponding to the read control data selected based on the determined priority order and are associated with an identical management ID to an order of receiving the read request data set.

Growing community of inventors

El Segundo, CA, United States of America

Dongpei Su