TWI763971B - Devices, systems and methods for personalized banner generation and display - Google Patents

Abstract

One or more computing devices, systems, and/or methods for personalized banner generation and display is provided. For example, an image comprising a product object depicting a product is identified. A background of the image is identified and removed. Features of the product object are evaluated to determine a product type of the product. A new background is selected based upon the product type. A banner comprising the new background and the product object positioned over the new background at a position within the banner is generated. The banner is rendered on a display of a computing device through a user interface.

Description

Apparatus, system and method for generating and displaying personalized advertising banners

Many users access content through computing devices, such as smart watches, mobile phones, computers, and/or many other electronic devices. Users can access displayed content through a user interface, such as a web browser that displays a website or an application that displays application content. When the user accesses the content through the user interface, additional disclosure and/or content of interest can also be effectively displayed to the user through the user interface. product information that may be of interest to readers. Since each user may have millions of different products and interests, it is not feasible to manually generate individual ad banners for each situation, product and user, and would consume a lot of computing resources and manual work to create and store Advertising banner.

In accordance with the present invention, one or more computing devices and/or methods are provided for personalized advertising banner generation and display. In a specific embodiment, it is determined whether the user accesses the content through the user interface displayed on the computing device. Information about the user, such as browsing history, purchase history, location, personal Personal setting information, age, gender, social network data, AD interaction data, email content, calendar data, etc., are used to determine the user's interest. Interest-related images can be obtained that are processed, such as through unsupervised image segmentation processing, to remove the background of the image and identify a product item within the image that describes the product. Other content such as logos, text or other objects are also removed from the product as part of removing the background. As such, the product item is isolated from further processing.

Features of the product item are evaluated to determine the product type of the product, eg, an image recognition network (eg, Image Recognition Network VGC-19) processes product item characteristics (eg, shape, color, and/or can be extracted from the image) other product identification features) to select the product type of the product from the product list. Choose a new context based on product type or manually assigned subject matter. Supplementary content (eg, product price, product description, etc.) and/or decorative items (eg, overlay images that can be overlaid with new backgrounds, such as sun items, cloud items, tree items, etc.) may be obtained by product type. In this way, an advertising banner containing the new background is generated. The product object, the supplemental content and/or the decorative object are inserted at selected positions of the advertising banner.

Apply visual adjustments to the advertising banner, such as modifying the brightness, saturation, size, position, color and/or other visual attributes of the new background, the product item, the decorative item and/or supplemental content. The advertising banner is generated mechanically by a computer in a way that is aesthetically pleasing and personal to the user. As such, the advertising banner is presented through the user interface on the computing device display.

100: Scheme

102: Services

104: Server

106: Local Area Network

108: WAN

110: Client Device

112: User

200: Schematic Architecture Diagram

202: Memory

204: Operating System

206: Server Application

208:Database

210: Processor

212: Communication bus

214: Network Adapter

216: Storage Components

218: Power Supply

220: Climate Control Unit

300: Schematic Architecture Diagram

301: Memory

302: User application

303: Operating System

304: Battery

306: Network Adapter

308: Display

310: Processor

311: Keyboard

312: Communication bus

318: Power Supply

319: GPS receiver

502: Swimming website

504: Recommended Mods

506: Product Type List

508: Detect

510: Identify

512: Image

514: Unsupervised Image Segmentation Processing

516: Remove

518: Guitar Product Object

520: Image Recognition Network

522: Judgment

524: Get

526: Construction

528: Advertising Banner

530: Transmission

600: Scheme

602: Non-transitory machine-readable media

604: The computer can read the data

606: read

608: Device

610: Reader

612: Processor executable instruction

616: Processor

While the techniques presented in this specification may be embodied in alternative forms, the specific specific embodiments illustrated in the accompanying drawings are merely examples of certain supplements to the illustrations provided in this specification. These specific examples should not be construed in a limiting manner, such as to limit the scope of the claims that follow.

The first figure is a schematic diagram of many network examples that can connect the server and the client.

The second figure is a schematic diagram of an example configuration of a server that may employ and/or implement at least a portion of the techniques presented in this specification.

Figure 3 is a schematic diagram of an example configuration of a client that may employ and/or implement at least a portion of the techniques presented in this specification.

The fourth figure is a flow diagram illustrating an example method of generating and displaying a personalized advertising banner.

Figure 5A is a block diagram of components illustrating an example system for personalized advertising banner generation and display, showing a user accessing a user interface.

Figure 5B is a component block diagram illustrating an example system for personalized advertising banner generation and display, showing an identifying image.

Figure 5 C is a component block diagram illustrating an example system for personalized advertising banner generation and display, showing removal of the image background.

Figure 5 D is a component block diagram illustrating an example system for personalized advertising banner generation and display, showing determining the product type of the product described by the product object.

Figure 5 E is a component block diagram illustrating an example system for personalized advertising banner generation and display, showing access to decoration modes and supplemental content.

Figure 5F is a component block diagram illustrating an example system for personalized advertising banner generation and display, showing the generation and presentation of an advertising banner on a display.

FIG. 6 is a schematic diagram of an example non-transitory machine-readable medium in accordance with one or more of the previously disclosed matters in this specification.

The present invention will be described more fully hereinafter with reference to the accompanying drawings which form a part of its subject matter, and which show by way of illustration specific example embodiments. This description is not intended to be an extensive or detailed discussion of known concepts. Details known to those skilled in the art may have been omitted, or may have been treated in a general manner.

The subject matter below can be embodied in various forms, such as methods, apparatus, components and/or systems. Therefore, this subject matter is not intended to be construed as limited to any example specific embodiments disclosed in this specification. Rather, example specific embodiments are provided for illustration only. This embodiment may take the form of, for example, hardware, software, firmware, or any combination of these.

1. Calculation scheme

The following provides a discussion of certain types of computing schemes that may utilize and/or implement the disclosed subject matter.

1.1. Networking

The first diagram is an interactive diagram illustrating a scheme 100 of a set of servers 104 providing a service 102 to a set of client devices 110 over many types of networks. Server 104 and/or client device 110 can send, receive, process and/or store many types of signals, such as in memory, as physical memory states.

The servers 104 of the service 102 may be internally connected through a local area network 106 (LAN), such as a wired network, wherein the network adapters on the individual servers 104 are connected through cables (eg, coaxial and/or fiber optic cables) Interconnected, and can be connected in various topologies (eg, busbars, command rings, grids, and/or trees). The servers 104 may be interconnected directly, or through one or more other network devices, such as routers, switches, and/or repeaters. The server 104 may utilize various physical network protocols (eg, Ethernet and/or Fibre Channel) and/or physical network protocols (eg, Internet Protocol (IP, Internet) Protocol) variants, Transmission Control Protocol (TCP, Transmission Control Protocol) and/or User Datagram Protocol (UDP, User Datagram Protocol). Local area network 106 may include, for example, analog telephone lines, such as twisted pair, coaxial cable, all or part of digital lines, including T1, T2, T3 or T4 type lines, Integrated Services Digital Network (ISDN) ), Digital Subscriber Line (DSL), wireless links including satellite links, or other communication links or channels, such as are well known to those skilled in the art. The local area network 106 may be based on one or more network architectures, such as master-slave, peer-to-peer, and/or mesh architectures, and/or various roles, such as management servers, authentication servers, security monitoring servers, for applications such as The files and database objects are organized architecturally by data storage devices, business logic servers, time synchronization servers, and/or front-end servers that provide a user-oriented interface to service 102 .

Likewise, local area network 106 may include one or more sub-networks, eg, may employ different architectures, may be compatible or compatible with different protocols, and/or may interoperate within local area network 106 . Additionally, various local area networks 106 may be interconnected; for example, routers may provide links between other separate and independent local area networks 106 .

In the scheme 100 of the first figure, the local area network 106 of the service 102 is connected to a wide area network (WAN) 108, which allows the service 102 to exchange data with other services 102 and/or client devices 110. Wide area network 108 may include many combinations of devices with varying degrees of dispersal and exposure, such as public wide area networks (eg, the Internet) and/or private networks (eg, virtual private networks (VPNs) for distributed enterprises network)).

In the scheme 100 of the first figure, the service 102 is accessible via the wide area network 108 by the users 112 of one or more client devices 110, such as portable media players (eg, e-book readers, audio devices or portable mobile gaming, sports or navigation devices), portable communication devices (such as cameras, phones, wearables or text chat devices), workstations and/or laptops. Individual client device 110 Communication with service 102 may be through a number of connections to wide area network 108 . As shown in the first such example, one or more client devices 110 may include cellular communication equipment and may communicate with service 102 by connecting to wide area network 108 via wireless area network 106 provided by a mobile communication provider. As shown in the second example, one or more client devices 110 may be provided via a wireless local area network 106 at a location, such as a user's home or workplace (eg, WiFi (Institute of Electrical and Electronics Engineers (IEEE) standard 802.11) network or Bluetooth (IEEE Std 802.15.1) Personal Area Network), utilizes connection to wide area network 108 to communicate with service 102. In this manner, the server 104 and the client device 110 can communicate over many types of networks. Other types of networks accessible by server 104 and/or client device 110 include mass storage devices such as network attached storage (NAS), storage area network (SAN), or others Computer or machine readable media form.

1.2. Server configuration

A second figure presents a schematic architectural diagram 200 of server 104 employing at least a portion of the techniques provided in this specification. The configuration or capability of this server 104, alone or in combination with other servers, to provide services such as service 102 varies widely.

Server 104 may include one or more processors 210 for processing instructions. The one or more processors 210 may optionally include multiple cores; one or more auxiliary processors, such as arithmetic auxiliary processors or integrated graphics processing units (GPUs); and/or one or more layers of local cache memory . The server 104 may include memory 202 in which a number of application programs are stored, such as an operating system 204; one or more server applications 206, such as a hypertext transfer protocol (HTTP) server, a file transfer protocol (FTP) server or Simple Mail Transfer Protocol (SMTP) servers; and/or various forms of data, such as database 208 or file systems. Server 104 may include various peripheral components, such as Wired and/or wireless network adapter 214, one or more storage components 216, such as hard drives, solid state storage devices (SSD), flash memory devices, and/or magnetic and/or optical disk readers.

The server 104 may include a motherboard having one or more communication buses 212 using various bus technologies, such as serial or variants of the AT Attachment (ATA) bus protocol, Universal Serial Bus (USB) Protocol and/or Small Computer System Interface (SCI) bus protocol to interconnect processor 210, memory 202 and many peripherals. In a multi-bus scheme, the communication bus 212 may interconnect the server 104 with at least one other server. Other components that may optionally include server 104 (although not shown in diagrammatic architecture diagram 200 of the second figure) include displays, display adapters such as graphics processing units (GPUs), input peripherals such as keyboards and/or A mouse, and a flash memory device, which can store basic input/output system (BIOS) routines that help boot the server 104 into a ready state.

The server 104 may operate within many physical enclosures, such as a desktop host, and/or may be integrated with a display as a "one-piece" device. Server 104 may be mounted horizontally and/or within a chassis or cabinet, and/or may simply comprise a set of interconnected components. Server 104 may include dedicated and/or shared power supplies 218 that supply and/or condition power to other components. Server 104 may provide power to and/or receive power from other servers and/or other devices. Server 104 may include dedicated and/or shared climate control units 220 that adjust climate attributes such as temperature, humidity, and/or airflow. Many such servers 104 can be configured and/or adapted to utilize at least a portion of the techniques provided in this specification.

1.3. Client device configuration

Figure 3 presents a schematic architectural diagram 300 of a client device 110 implementing at least a portion of the techniques provided in this specification. The client device 110 may vary widely in configuration or capability to provide various functions to users, such as user 112 . Client device 110 may be provided in a variety of form factors, such as a desktop or tower workstation; an "all-in-one" device with integrated display 308; Laptops, Tablets, Convertible Tablets, or Palmtops; Mounted in Headphones, Glasses, Headphones, and/or Watches, and/or Integrated into Clothing, and/or Parts of Furniture, such as Tabletops, and/or or other devices, such as wearables in vehicles or homes. The client device 110 may serve users in various roles, such as workstations, kiosks, media players, gaming devices, and/or appliances.

The client device 110 may include one or more processors 310 for processing instructions. The one or more processors 310 may optionally include multiple cores, one or more auxiliary processors, such as arithmetic auxiliary processors or integrated graphics processing units (GPUs), and/or one or more layers of local cache memory. Client device 110 may include memory 301 that stores many types of applications such as operating system 303, such as file applications, media applications, file and/or data access applications, such as web browsers and/or email One or more user applications 302 such as communication applications such as clients, utilities and/or games, and/or various peripheral drivers. The client device 110 may include various peripheral components, such as a wired and/or wireless network adapter 306 that can be connected to a local area network and/or a wide area network; one or more output components, such as a display adapter ( Display 308, which may optionally include a graphics processing unit (GPU), a sound adapter coupled to speakers, and/or a printer; input devices for receiving input from a user, such as keyboard 311, mouse, microphone , touch-sensitive components of camera and/or display 308; and/or environmental sensors, such as global positioning system (GPS) receiver 319, compass, accelerometer that detects the position, velocity, and/or acceleration of client device 110 and/or a gyroscope for detecting the physical orientation of the client device 110 . Other components that may optionally be included in the client device 110 (although not shown in the diagrammatic architecture diagram 300 of the third figure) include one or more storage components, such as hard drives, solid state storage devices (SSD), fast A flash memory device and/or a disk and/or optical disk reader; and/or a flash memory device that can store basic input/output system (BIOS) routines to help boot the client device 104 into a ready state; and climate control units that regulate climate properties such as temperature, humidity and airflow.

The client device 110 may include a motherboard having one or more communication buses 312 using various bus technologies, such as serial or parallel AT Attachment (ATA) bus protocol variants, generic serial The bus (USB) protocol and/or the Small Computer System Interface (SCI) bus protocol are used to interconnect the processor 310, the memory 301 and many peripherals. Client device 110 may include dedicated and/or shared power supplies 318 that supply and/or regulate power for other components, and/or store power for use when client device 110 is not connected to a power source through power supply 318 battery 304. The client device 110 may provide power to and/or receive power from other client devices.

In some aspects, when the user 112 interacts with a software application (eg, an instant messenger and/or an email application) on the client device 110, the descriptive content under the signal type or the in-memory Stored entity status (eg, email address, instant messenger ID, phone number, zip code, message content, date and/or time). Descriptive content is often stored along with contextual content, for example, the source of a phone number (eg, a communication received from another user via an instant messenger application) may be stored as contextual content associated with the phone number. Thus, contextual content can identify the circumstances surrounding when the phone number was received (eg, the date or time the phone number was received) and can be associated with descriptive content. The contextual content can be used, for example, for subsequent searches for relevant descriptive content, eg, a search can be initiated for a phone number received from a particular individual, received through an instant messenger application, or received on a known date or time. Client device 110 may include one or more servers that may locally serve client device 110 and/or other client devices of user 112 and/or other individuals. For example, a locally installed web server may serve web page content along with network requests submitted locally. Many such client devices 110 may be configured and/or adapted to employ at least a portion of the techniques provided in this specification.

2. Technology presented

The present disclosure provides one or more systems and/or techniques for the generation and display of personalized advertising banners. In particular, the present system provides machine advertising banner design through a computer capable of automatically designing and generating personalized advertising banners (eg, dynamically changing as the user accesses the user interface that will display the advertising banner through the user interface). Personalized advertising banners provide an improved user experience, as users will be more likely to interact with advertising banners that are personalized to user interests and design tastes. Otherwise, for people to use computers to generate millions of advertising banners and then personalize them for each possible situation, for example due to millions of different products and users with different tastes and interests, such as price or quantity of these Product information changes rapidly, etc., which would be excessively manual and computer resource intensive.

As will be discussed further, the present system is configured to perform background removal of imagery depicting the product, identifying suitable new background candidates and decorators, arranging and constructing advertising banners based on the product type of the product, and applying a color scheme to the advertising banner, to automatically generate personalized advertising banners. The system is configured to generate personalized advertising banners in real time when the user accesses the user interface (eg, browses to a website). The system is set up to automatically collect a large number of layers for machine learning training on how to design and generate advertising banners. The system is set up to build an ecosystem (repository) of personalized ad banners, so user feedback (eg, interaction with ad banners or user ignoring ad banners) is used to train and modify recommendations for designing and generating ad banners module.

The system is configured to automatically resize and/or automatically crop personalized advertising banners based on user browsing behavior and/or device properties (eg, screen size, viewing orientation, resolution, etc.). The present system is configured to automatically generate and personalize ad banners for content campaign execution as needed (eg, each action for generating and personalizing ad banners can be turned on or off as needed). The system is configured to generate ad banner configurations based on user scenarios, such as using different ad banner configurations based on the location within the user interface on which the ad banner is to be displayed. The system is configured to generate based on user feedback Advertising banner configuration, for example, using different advertising banner configurations for different users based on the user's past reactions to advertising banners.

The system is configured to perform background removal by detecting main items (eg, product objects) within the image to define the background area to be removed from the image. Areas within the image are segmented based on saturation and brightness. Background areas are removed based on item detection and segmented areas, while leaving the area containing the main item for further processing.

The system is configured to apply a color scheme to an advertising banner. In particular, items, backgrounds, decorative patterns, text and/or other elements (objects) are separated and colored according to different shades on the color wheel. The system is arranged to aesthetically determine where a host item (eg, a product item) is to be positioned in the advertising banner based on the geometric center of the item. The system is set up to aesthetically disperse elements on the advertising banner. The system is set up to build and review ad banners through user feedback to enhance the aesthetics produced by the ad banner using machine learning to process user feedback.

Through the exemplary method 400 in FIG. 4, combined with the descriptions in FIG. 5A to FIG. 5F, a specific embodiment of generating and displaying a personalized advertisement banner is illustrated. In an example, a user may use a computing device, such as a mobile device, computer, wearable device, smart watch, etc., to access a user interface containing swimming team website content, as shown in FIG. 5A. In one embodiment, a recommendation module 504 (eg, hardware and/or software executing one or more computers for designing, generating, and displaying/rendering personalized advertising banners) may detect 508 user memory. Take this user interface. In one embodiment, the user is identified, if possible, to obtain and evaluate user information to determine the user's interests (eg, access to music website browsing history, social networking sites to recommend guitar purchases) Road posts, emails from music stores, calendar entries about band practice, user profile information such as age and gender, etc.). For example, the user information may indicate that the user is interested in guitars.

The recommendation module 504 is configured to identify 510 the image 512 describing the product, as illustrated in Figure 5B. For example, the recommendation module 504 identifies 510 the image 512 based on the image 512 describing the guitar and the user is interested in the guitar. Imagery 512 may be identified from a library of images describing the product or from any other source of content. In this way, the image can be identified according to the image related to the user's interest.

The imagery 512 is processed to generate personalized advertising banners that a particular user may find beautiful and include relevant content, thereby increasing the likelihood that the user will want to interact with the advertising banner.

At step 402, the background of the image 512 has been removed 516, as illustrated in the fifth panel C. In one embodiment, the recommendation module 504 performs an unsupervised image segmentation process 514 to distinguish a first portion of the image 512 as describing a product, such as the guitar product object 518 described within the image 512, and a second portion of the image 512 to describe the product Sections are divided into descriptive backgrounds (eg, sun objects, tree objects, sky objects, grass objects, cloud objects, and/or objects other than the guitar product object 518). In particular, the unsupervised image segmentation process 514 performs edge detection to detect the edges of objects depicted in the image 512, such as cloud objects, guitar product objects 518, sun objects, and the like. The unsupervised image segmentation process 514 calculates the area of each object, eg, the number of pixels representing the pixels of each object. In some examples, an image of a product may describe the product at a larger size than other objects depicted in the image, guitar product item 518 is identified as a guitar product item based on, for example, a guitar product item having a larger area than other objects depicted in image 512 518 to describe a product. It will be appreciated that various other techniques may be used to detect image regions used to describe products (eg, machine learning models may be used for image recognition such as feature extraction and entity recognition). In a specific embodiment, multiple product items may be identified if there is a set of items (eg, images depicting multiple products) that have a relatively larger area than other items.

Objects other than guitar product objects 518 are removed from image 512, eg, sun objects, cloud objects, grass objects, tree objects, etc. are removed as part of removing 516 the background. An object is removed by assigning the average color of the surrounding pixels as the pixel from which the object was removed.

At step 404, the characteristics of the guitar product item 518 are evaluated to determine 522 the product type of the product, as illustrated in the fifth D diagram. In particular, the recommendation module 504 maintains a product type list 506 for various product types and/or indicates characteristics of this product type (eg, cyberpunk, tech, sweet pink, or any other type of product). These features can correspond to shape, size, color, and/or various other features that can be captured from the image for detecting the product type of the product. In an example, image recognition network 520 (eg, image recognition network VGC-19) or other machine learning techniques are used to process the features extracted from image 512 to identify corresponding ones from product type list 506 of guitar product object 518 product type, such as music product type.

This product type can be used to select a new background, and to obtain 524 decorative patterns and supplemental content for use in constructing a personalized advertising banner to display to the user, as illustrated in Figure 5E. In particular embodiments, the new context is selected based on manually assignable subject matter. In the specific embodiment of acquiring 524 the supplementary content, according to how much each supplementary content corresponds to the product title of the guitar product or other identification information of the guitar product, a term frequency-inverse document frequency (term frequency-inverse document frequency) is performed. algorithm) to calculate scores for a plurality of supplemental content (eg, images, text, audio, video, product pricing, number of products available, product descriptions, keywords about products and/or available from various content sources, such as supplemental content repositories, websites, web services, etc., other content obtained). As such, one or more supplemental content items are selected based on one or more supplemental content items having scores above a threshold (eg, the highest score). For example, get the short message "Axe Guitar!!" from a website and use it as a supplement to generate an ad banner.

In particular embodiments of obtaining trim patterns, from a data source (eg, a trim pattern repository, website, web service, etc.) based on one or more trim patterns corresponding to the product type or other identifying information for the guitar product Select one or more decoration modes. For example, get 524 musical note objects from a data source to be used as decoration patterns to generate advertising banners. In a particular embodiment, at step 406, a new background is selected according to the product type of the guitar product. The new background can include shapes, patterns, drawings, colors, videos, images, or other content related to the type of music product.

At step 408, an advertisement banner 528 is constructed at step 526 using the new background, guitar product object 518, and/or any decorative patterns or supplemental content acquired for the advertisement banner 528, as illustrated in FIG. 5F. In one embodiment, the configuration of the advertising banner 528 is determined, such as the guitar product item 518 to be inserted within the advertising banner 528, a decorative pattern such as a musical note item, and/or supplemental content such as the text message "Axe Guitar!!" size and location. This configuration may be based on where within the swimming website 502 the advertising banner 528 will be displayed. The configuration may be based on previous user interactions with the content, such as the user previously interacted with an ad banner where the product item was in the middle of the ad banner and had an oversized size, and the user previously ignored the product item that was not in the middle of the ad banner and had a smaller size product item Advertising banner. Thus, the guitar product item 518 can be placed in the middle of the advertising banner 528 and the size of the guitar product item 518 can be increased. As such, the user may be more able to interact with the advertising banner 528 because the configuration of the advertising banner 528 is personalized to the user. In one embodiment, the ad banner 528 is automatically sized and/or positioned within the swimming website 502 based on the user's browsing behavior (eg, the user does not typically scroll down the website, so the ad banner 528 The ad banner 528 may be positioned toward the top of the swimming website 502 to draw the user's attention to the ad banner 528; larger sized ad banners are typically ignored by the user, so the ad banner 528 size may be reduced, etc.).

In particular embodiments of constructing 526 ad banner 528, visual adjustments are applied to ad banner 528, such as adjusting saturation, brightness, or other visual characteristics (eg, color, hue, size, etc.). Visual adjustments can be applied to new backgrounds, guitar product items 518, decoration modes, supplemental content, and more. In an example, the dominant color of guitar product object 518 is detected. The dominant color is transferred from the Red, Green, Blue (RGB) color space to the Hue, Saturation, Value (HSV) color space. Complementary and/or analogous colors of the primary color are identified using the HSV color space (eg, the complementary color is within 180 degrees of the primary color and the analogous color is within 120 degrees of the primary color). Visual adjustments are applied to the ad banner 528 based on complementary and/or similar colors, such as by recoloring objects, decorative patterns, supplemental content, and/or new backgrounds based on color.

The advertising banner 528 is transmitted 30, for example, over a network to the user's computing device for presentation through the swimming website 502 on the display of the computing device. In one embodiment, the user's interaction with the ad banner 528 through the swim website 502 is monitored, such as whether the user clicks on the ad banner 528, scrolls quickly through the ad banner 528, scrolls to highlight/see the ad banner 528 through the swim website 502 s position. Recommendation module 504, such as various machine learning techniques implemented by recommendation module 504 to select new backgrounds, generate ad banners 528, visually adjust fit to ad banners 528, select decorative patterns or supplemental content, identify product types, etc., depending on usage interact with the user to update. For example, the parameters and/or other logic of the recommendation module 504 may be adjusted based on user interaction (eg, the higher values of the parameters used to create the ad banner 528 may be maintained or weighted based on user interaction with the ad banner 528, Or the value may be changed or weighted lower based on the user not interacting with the ad banner 528). In this way, the creation of personalized advertising banners is improved.

The sixth figure is a diagram of an approach 600 for an example non-transitory machine-readable medium 602. Non-transitory machine-readable medium 602 can include processor-executable instructions 612 that, when executed by processor 616, cause (eg, by processor 616) to perform the performance of at least some of this specification. The non-transitory machine-readable medium 602 may include memory semiconductors (eg, using static random access memory) (SRAM), Dynamic Random Access Memory (DRAM) and/or Synchronous Dynamic Random Access Memory (SDRAM) technology semiconductors), hard drives, flash memory devices, or magnetic or optical disks (such as CDs, digital CD-ROM (DVD) or Floppy Disk). An example non-transitory machine-readable medium 602 stores computer-readable data 604 that, when read by a reader 610 of a device 608 (eg, a read head of a hard drive or a read operation invoked on a solid-state storage device) At 606, the expression processor can execute the instructions 612. In some embodiments, processor-executable instructions 612, when executed, result in the performance of operations, such as at least some of the example methods 400 of FIG. 4 . In some embodiments, the processor-executable instructions 612 are arranged to cause a system, such as at least some of the example systems 500 of FIGS. 5A-5F, to be implemented.

3. Glossary usage

As used in this application, "component," "module," "system," "interface," etc. are used to mean computer-related entities, hardware, a combination of hardware and software, software, or software in execution . For example, a component can be, but is not limited to, a process executing on a processor, a processor, an object, an executable, a thread, a program, and/or a computer. For example, both an application executing on a controller and the controller can be one component. There may be one or more components within a process and/or thread, and a component may be localized on a computer and/or distributed between two or more computers.

Unless stated otherwise, "first," "second," and/or similar ordinals are not meant to imply temporality, spatiality, ordering, or the like. Rather, these terms are used only as identifiers, names, etc. of features, elements, items, and the like. For example, a first object and a second object generally correspond to object A and object B, or to two different or two identical objects or the same object.

Furthermore, "example" as used in this specification means an example, instance, illustration, etc., and is not necessarily advantageous. As used in this specification, "or" means an inclusive "or" rather than an exclusive "or." In addition, if the word "a" is used in this application, unless there is a special indication or the content is clearly displayed as a single form, otherwise it should generally be regarded as "one or more". Furthermore, at least one of A and B, etc. generally means A or B or both A and B. In addition, the use of "comprising", "having", "possessing", "containing" and/or related variations thereof in the embodiments or the scope of the patent application is similar to the usage of the word "including".

Although the claimed subject matter is described in terms of specific structural features and/or methodological acts, it should be understood that the subject matter defined in the following claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed merely as example forms for implementing at least some of the claimed scope.

Further, the claimed subject matter can be implemented as a method, apparatus or article of manufacture using standard programming and/or engineering techniques to produce software, firmware, hardware, or any combination of these, to control a computer to implement the disclosed subject matter. The term "article of manufacture" as used in this specification is meant to encompass a computer program accessible from any computer-readable device, carrier or media. Of course, many modifications may be made to this configuration without departing from the scope or spirit of the claimed subject matter.

The operation of many specific embodiments is provided in this specification. In one embodiment, one or more of the operations described may constitute computer-readable instructions stored on one or more computer-readable media that, if executed by a computing device, cause the computing device to execute the described operation. The order in which some or all operations are described should not be construed as implying that the operations must be in that order, alternative orders will be apparent to those skilled in the art having the benefit of this description. Furthermore, it should be understood that not all operations are necessarily present in every specific example provided in this specification. Furthermore, it should be understood that not all operations are necessarily available in some specific embodiments.

Furthermore, while the invention has been shown and described in one or more implementations, equivalent changes and modifications may occur to those skilled in the art upon reading and understanding the specification and drawings. The present invention includes all such modifications and changes, and is limited only by the scope of the following claims. Especially with regard to the above groups The various functions performed by a component (eg, an element, resource, etc.), unless otherwise indicated, the terms used to describe such a component correspond to any component that performs the specified function (eg, equivalent function) of the described component, even if not structurally equivalent on the disclosed structure. Furthermore, although a particular function of the present invention is disclosed with respect to only one of several implementations, this function may be combined with one or more other features of other implementations to be desired and advantageous for any known or particular application.

500: System

502: Swimming website

504: Recommended Mods

506: Product Type List

526: Construction

528: Advertising Banner

530: Transmission

Claims (20)

A method for generating and displaying personalized advertising banners, comprising: on a processor of a computing device, executing instructions to cause the computing device to perform operations, the operations comprising: identifying an image comprising one or more objects; Distinguishing a first portion of the image of a product object corresponding to one or more objects from a second portion of the image corresponding to a background comprising at least one of the one or more objects different from the product object an object; removing the background of the image; evaluating characteristics of the product object to determine the product type of a product described by the product object; selecting a new background based on the product type; and generating an advertisement that includes the new background banner, and the product object is placed on the new background at a location within the advertising banner. The method of claim 1, wherein the distinguishing is accomplished by performing an unsupervised image segmentation process. The method of claim 2, wherein the performing an unsupervised image segmentation process comprises: detecting edges of objects described in the image; and calculating the area of each object. The method of claim 1, wherein the removing the background comprises: assigning an average color of surrounding pixels as pixels of the background. The method as claimed in claim 1, wherein the evaluating the features of the product object comprises: extracting the features of the product object; and using the features to select the product type from a plurality of product types. The method of claim 1 of the patent scope, wherein the evaluating the feature of the product object comprises: using an image recognition network to process the image to determine the product type. The method of claim 1 of the scope of application includes: selecting one or more decorative pattern objects according to the product type; and inserting the one or more decorative pattern objects into the advertisement banner. The method of claim 7, comprising: applying a visual adjustment to the advertising banner, wherein the visual adjustment modifies at least one of a saturation, a brightness, or a visual characteristic of the one or more decorative pattern objects . The method of claim 1, comprising: applying a visual adjustment to the advertising banner, wherein the visual adjustment modifies at least one of a saturation, a brightness, or a visual characteristic of the new background. The method of claim 1 of the scope of the application includes: detecting a dominant color of the product object; transferring the dominant color from a red, green, blue (RGB) color space to a color, saturation, value (HSV) color space ) color space; use the HSV color space to identify complementary and analogous colors of the primary color; and apply a visual adjustment to the advertising banner based on the complementary and analogous colors. A computing device for personalized advertising banner generation and display, comprising: a processor; and memory including processor-executable instructions that, when executed by the processor, cause operations to be performed, the operations including: identifying an image containing one or more objects; distinguishing objects corresponding to one or more A first portion of the image of a product object of an object and a second portion of the image corresponding to a background containing at least one object from one or more objects different from the product object; removing the the background of the image; evaluating characteristics of the product object to determine the product type of a product the product object describes; selecting a new background based on the product type; and generating an advertising banner that includes the new background, and the product object placed on the new background at a location within the ad banner. The computing device of claim 11, wherein the operations include: obtaining supplementary content about the product; and inserting the supplementary content into the advertising banner. The computing device of claim 12, wherein the operations include: executing a word frequency-reverse file frequency algorithm to calculate the plurality of supplementary contents according to the corresponding relationship between the plurality of supplementary contents and a product title of the product a score of the supplementary content; and selecting the supplementary content from among the plurality of supplementary content according to the score of the supplementary content exceeding a threshold. The computing device of claim 12, wherein the supplementary content includes at least one of product features, prices, or keywords related to the product. The computing device of claim 11, wherein the advertisement banner is generated through a recommendation model, and wherein the operations include: monitoring user interaction with the advertisement banner; and updating the recommendation model according to the user interaction. The computing device of claim 15, wherein one or more parameters of the recommendation model are updated according to user interaction, and wherein the one or more parameters are used to select the new background, generate the advertising banner, At least one of applying a visual adjustment to the advertising banner, or selecting decorative mode objects to be included in the advertising banner. A non-transitory machine-readable medium for the generation and display of personalized advertising banners having stored thereon processor-executable instructions that, when executed, cause operations to be performed, the operations comprising: removing a image background of a product object; evaluating characteristics of a product object to determine the product type of the product object; selecting a new background and one or more decorative pattern objects based on the product type; and generating an advertising banner including the A new background and the product object located on the new background within the advertising banner, at least one of the following: removing a background, including: performing an unsupervised image segmentation process to distinguish one of the images describing the product The first part and a second part of the image describing the background, wherein the performing an unsupervised image segmentation process includes: detecting edges of objects described in the image; calculating the area of each object; Identifying the product object describing the product according to the product object having an area larger than other objects; and removing the objects other than the product object from the image; the operation comprises: detecting the presence of the product object a dominant color; transfer the dominant color from a red, green, blue (RGB) color space to a hue, saturation, value (HSV) color space; use the HSV color space to identify the complementary color of the dominant color and similar color; and applying a visual adjustment to the advertising banner based on the complementary color and the analogous color; or the operation includes: executing a word frequency-reverse file frequency algorithm to provide a product title for the product based on a plurality of supplemental content and the product According to the corresponding relationship between the plurality of supplementary contents, the scores of the plurality of supplementary contents are calculated; according to the score of the supplementary contents exceeding a threshold, the supplementary contents are selected from the plurality of supplementary contents; and the supplementary contents are inserted into the advertisement banner. The non-transitory machine-readable medium of claim 17, wherein the operations include: at least one of a property of a device or a user browsing behavior of a user using the device to access a user interface , which automatically performs at least one of resizing or automatic cropping of the ad banner. The non-transitory machine-readable medium of claim 17, wherein the operations include: determining the product object in the advertising banner and the one or the other according to the position of the advertising banner in a user interface Configuration of multiple decoration mode objects. The non-transitory machine-readable medium of claim 17, wherein the operations include: determining the product object and the one or more decoration patterns in the advertising banner based on previous user interactions with the content Configuration of objects.

Images