A method is disclosed. The method includes receiving print job data at a printing system, tracking resource objects in print job data while the print job data is rasterized to generate rasterized print data and updating a count of resource objects for print job data confirmed as printed.

A print control device includes a control unit configured to acquire print image data from a storage device according to a print request and to cause a printer 100 which is configured to execute printing based on the print image data to execute the printing, wherein the control unit acquires from the storage device, operation status information of the storage device storing the print image data and attribute information of the print image data stored in the storage device and enables the operation status information and the attribute information to be displayed on a display unit.

An information processing apparatus includes an input unit, a display unit, a designation unit and an update unit. The input unit inputs a plurality of parts configuring one character. The display unit displays the plurality of parts. The designation unit designates one or more parts selected from the plurality of parts by a user. The update unit updates a display on the display unit such that the designated one or more parts are deleted from the one character and the other parts are remained displayed on the display unit.

A printing system is disclosed. The printing system includes a color management unit having a hash table to store input colors and corresponding output colors and a lookup table (LUT). The printing system also includes one or more processors to reinitialize the hash table based on a hash table time and a interpolation time derived from concurrent real time measurements.

Embodiments described herein provide for the reuse of some locations in a hash table of color conversions during processing of a print job, while preventing the reuse of other locations. If a determination is made that the hash table is to be re-initialized, then the locations in the hash table that are marked as non-persistent are marked as saved to allow them to be reused, while the locations in the hash table that are marked as persistent are prevented from being reused. During processing of the print job subsequent to re-initialization of the hash table, if a color in the print job does not have a color conversion, then a location in the hash table is identified that is marked as saved. If the location stores the color conversion for the color, then the non-persistent mark is restored for the location to bypass performing a color conversion for the color.

An information processing apparatus includes a storage unit and a processor. A font file includes type face data and serif data. When a sans-serif font is selected, the processor generates sans-serif character data on the basis of the type face data. When a serif font selected, the processor generates sans-serif character data, generates serif image data, and adds generated the serif image data to generated the sans-serif character data to generate serif character data.

In implementations of systems for font replacement in print workflows, a computing device implements a print system to receive print request data describing a document having a corpus of text rendered using a font that is not available to the print system. The print system extracts the corpus of text from the document and generates an indication of a context category of the corpus of text using a machine learning model training to classify context categories of text inputs using training data describing a different corpus of text. A replacement font is identified based on the indication of the context category from replacement font data describing a plurality of candidate replacement fonts. The printing system generates a raster image depicting the corpus of text rendered using the replacement font.

A method and system to generate training data for a deep learning model in memory instead of loading pre-generated data from disk storage. A corpus may be stored as lines of text. The lines of text can be manipulated in the memory of a central processing unit (CPU) of a computing system, using asynchronous multi-processing, in parallel with a training process being conducted on the system's graphics processing unit (GPU). With such an approach, for a given line of text, it is possible to take advantage of different fonts and different types of image augmentation without having to put the images in disk storage for subsequent retrieval. Consequently, the same line of text can be used to generate different training images for use in different epochs, providing more variability in training data (no training sample is trained on more than once). A single training corpus may yield many different training data sets. In one aspect, the model being trained is a deep learning model, which may be one of several different types of neural networks. The training enables the deep learning model to perform OCR on line images.

In implementations of systems for font replacement in print workflows, a computing device implements a print system to receive print request data describing a document having a corpus of text rendered using a font that is not available to the print system. The print system extracts the corpus of text from the document and generates an indication of a context category of the corpus of text using a machine learning model training to classify context categories of text inputs using training data describing a different corpus of text. A replacement font is identified based on the indication of the context category from replacement font data describing a plurality of candidate replacement fonts. The printing system generates a raster image depicting the corpus of text rendered using the replacement font.

A method and system to generate training data for a deep learning model in memory instead of loading pre-generated data from disk storage. A corpus may be stored as lines of text. The lines of text can be manipulated in the memory of a central processing unit (CPU) of a computing system, using asynchronous multi-processing, in parallel with a training process being conducted on the system's graphics processing unit (GPU). With such an approach, for a given line of text, it is possible to take advantage of different fonts and different types of image augmentation without having to put the images in disk storage for subsequent retrieval. Consequently, the same line of text can be used to generate different training images for use in different epochs, providing more variability in training data (no training sample is trained on more than once). A single training corpus may yield many different training data sets. In one aspect, the model being trained is a deep learning model, which may be one of several different types of neural networks. The training enables the deep learning model to perform OCR on line images.