Accelerating your advertising and marketing ideation with generative AI – Half 2: Generate customized advertising and marketing pictures from historic references

Advertising groups face main challenges creating campaigns in immediately’s digital atmosphere. They need to navigate via advanced information analytics and quickly altering client preferences to provide partaking, customized content material throughout a number of channels whereas sustaining model consistency and dealing inside tight deadlines. Utilizing generative AI can streamline and speed up the inventive course of whereas sustaining alignment with enterprise aims. Certainly, in response to McKinsey’s “The State of AI in 2023” report, 72% of organizations now combine AI into their operations, with advertising and marketing rising as a key space of implementation.

Constructing upon our earlier work of advertising and marketing marketing campaign picture era utilizing Amazon Nova basis fashions, on this put up, we reveal methods to improve picture era by studying from earlier advertising and marketing campaigns. We discover methods to combine Amazon Bedrock, AWS Lambda, and Amazon OpenSearch Serverless to create a sophisticated picture era system that makes use of reference campaigns to keep up model tips, ship constant content material, and improve the effectiveness and effectivity of recent marketing campaign creation.

The worth of earlier marketing campaign info

Historic marketing campaign information serves as a strong basis for creating efficient advertising and marketing content material. By analyzing efficiency patterns throughout previous campaigns, groups can establish and replicate profitable inventive components that persistently drive greater engagement charges and conversions. These patterns would possibly embody particular coloration schemes, picture compositions, or visible storytelling methods that resonate with goal audiences. Earlier marketing campaign property additionally function confirmed references for sustaining constant model voice and visible id throughout channels. This consistency is essential for constructing model recognition and belief, particularly in multi-channel advertising and marketing environments the place coherent messaging is crucial.

On this put up, we discover methods to use historic marketing campaign property in advertising and marketing content material creation. We enrich reference pictures with useful metadata, together with marketing campaign particulars and AI-generated picture descriptions, and course of them via embedding fashions. By integrating these reference property with AI-powered content material era, advertising and marketing groups can rework previous successes into actionable insights for future campaigns. Organizations can use this data-driven method to scale their advertising and marketing efforts whereas sustaining high quality and consistency, leading to extra environment friendly useful resource utilization and improved marketing campaign efficiency. We’ll reveal how this systematic technique of utilizing earlier marketing campaign information can considerably improve advertising and marketing methods and outcomes.

Resolution overview

In our earlier put up, we applied a advertising and marketing marketing campaign picture generator utilizing Amazon Nova Professional and Amazon Nova Canvas. On this put up, we discover methods to improve this answer by incorporating a reference picture search engine that makes use of historic marketing campaign property to enhance era outcomes. The next structure diagram illustrates the answer:

The primary structure elements are defined within the following checklist:

1. Our system begins with a web-based UI that customers can entry to begin the creation of recent advertising and marketing marketing campaign pictures. Amazon Cognito handles consumer authentication and administration, serving to to make sure safe entry to the platform.
2. The historic advertising and marketing property are uploaded to Amazon Easy Storage Service (Amazon S3) to construct a related reference library. This add course of is initiated via Amazon API Gateway. On this put up, we use the publicly out there COCO (Frequent Objects in Context) dataset as our supply of reference pictures.
3. The picture processing AWS Step Features workflow is triggered via API Gateway and processes pictures in three steps:
   1. A Lambda perform (DescribeImgFunction) makes use of the Amazon Nova Professional mannequin to explain the photographs and establish their key components.
   2. A Lambda perform (EmbedImgFunction) transforms the photographs into embeddings utilizing the Amazon Titan Multimodal Embeddings basis mannequin.
   3. A Lambda perform (IndexDataFunction) shops the reference picture embeddings in an OpenSearch Serverless index, enabling fast similarity searches.
4. This step bridges asset discovery and content material era. When customers provoke a brand new marketing campaign, a Lambda perform (GenerateRecommendationsFunction) transforms the marketing campaign necessities into vector embeddings and performs a similarity search within the OpenSearch Serverless index to establish essentially the most related reference pictures. The descriptions of chosen reference pictures are then integrated into an enhanced immediate via a Lambda perform (GeneratePromptFunction). This immediate powers the creation of recent marketing campaign pictures utilizing Amazon Bedrock via a Lambda perform (GenerateNewImagesFunction). For detailed details about the picture era course of, see our earlier weblog.

Our answer is out there in GitHub. To deploy this challenge, observe the directions out there within the README file.

Process

On this part, we look at the technical elements of our answer, from reference picture processing via last advertising and marketing content material era.

Analyzing the reference picture dataset

Step one in our AWS Step Features workflow is analyzing reference pictures utilizing the Lambda Perform DescribeImgFunction. This useful resource makes use of Amazon Nova Professional 1.0 to generate two key elements for every picture: an in depth description and a listing of components current within the picture. These metadata elements might be built-in into our vector database index later and used for creating new marketing campaign visuals.

For implementation particulars, together with the whole immediate template and Lambda perform code, see our GitHub repository. The next is the structured output generated by the perform when introduced with a picture:

{
  "statusCode": 201,
  "physique": {
    "labels_list": [
      "baseball player in white t-shirt",
      "baseball player in green t-shirt",
      "blue helmet",
      "green cap",
      "baseball glove",
      "baseball field",
      "trees",
      "grass"
    ],
    "description": "A picture exhibiting two folks taking part in baseball. The individual in entrance, sporting a white t-shirt and blue helmet, is working in direction of the bottom. The individual behind, sporting a inexperienced t-shirt and inexperienced cap, is holding a baseball glove in his proper hand, presumably making ready to catch the ball. The background features a lush inexperienced space with bushes and a dust baseball subject.",
    "msg": "success"
  }
}

Producing reference picture embeddings

The Lambda perform EmbedImgFunction encodes the reference pictures into vector representations utilizing the Amazon Titan Multimodal Embeddings mannequin. This mannequin can embed each modalities right into a joint house the place textual content and pictures are represented as numerical vectors in the identical dimensional house. On this unified illustration, semantically comparable objects (whether or not textual content or pictures) are positioned nearer collectively. The mannequin preserves semantic relationships inside and throughout modalities, enabling direct comparisons between any mixture of pictures and textual content. This permits highly effective capabilities corresponding to text-based picture search, picture similarity search, and mixed textual content and picture search.

The next code demonstrates the important logic for changing pictures into vector embeddings. For the whole implementation of the Lambda perform, see our GitHub repository.

with open(image_path, "rb") as image_file:
    input_image = base64.b64encode(image_file.learn()).decode('utf8')

response = bedrock_runtime.invoke_model(
    physique=json.dumps({
        "inputImage": input_image,
        "embeddingConfig": {
            "outputEmbeddingLength": dimension
        }
    }),
    modelId=model_id
)
json.masses(response.get("physique").learn())

The perform outputs a structured response containing the picture particulars and its embedding vector, as proven within the following instance.

{
    'filename': '000000000872.jpg',
    'file_path': '{AMAZON_S3_PATH}',
    'embedding': [
        0.040705927,
        -0.007597826,
        -0.013537944,
        -0.038679842,
        ... // 1,024-dimensional vector by default, though this can be adjusted
    ]
}

Index reference pictures with Amazon Bedrock and OpenSearch Serverless

Our answer makes use of OpenSearch Serverless to allow environment friendly vector search capabilities for reference pictures. This course of includes two most important steps: establishing the search infrastructure after which populating it with reference picture information.

Creation of the search index

Earlier than indexing our reference pictures, we have to arrange the suitable search infrastructure. When our stack is deployed, it provisions a vector search assortment in OpenSearch Serverless, which robotically handles scaling and infrastructure administration. Inside this assortment, we create a search index utilizing the Lambda perform CreateOpenSearchIndexFn.

Our index mappings configuration, proven within the following code, defines the vector similarity algorithm and the marketing campaign metadata fields for filtering. We use the Hierarchical Navigable Small World (HNSW) algorithm, offering an optimum steadiness between search pace and accuracy. The marketing campaign metadata consists of an goal subject that captures marketing campaign objectives (corresponding to clicks, consciousness, or likes) and a node subject that identifies goal audiences (corresponding to followers, clients, or new clients). By filtering search outcomes utilizing these fields, we will help be sure that reference pictures come from campaigns with matching aims and goal audiences, sustaining alignment in our advertising and marketing method.

{
    "mappings": {
        "properties": {
            "outcomes": {"sort": "float"},
            "node": {"sort": "key phrase"},
            "goal": {"sort": "key phrase"},
            "image_s3_uri": {"sort": "textual content"},
            "image_description": {"sort": "textual content"},
            "img_element_list": {"sort": "textual content"},
            "embeddings": {
                "sort": "knn_vector",
                "dimension": 1024,
                "technique": {
                    "engine": "nmslib",
                    "space_type": "cosinesimil",
                    "identify": "hnsw",
                    "parameters": {"ef_construction": 512, "m": 16}
                }
            }
        }
    }
}

For the whole implementation particulars, together with index settings and extra configurations, see our GitHub repository.

Indexing reference pictures

With our search index in place, we are able to now populate it with reference picture information. The Lambda perform IndexDataFunction handles this course of by connecting to the OpenSearch Serverless index and storing every picture’s vector embedding alongside its metadata (marketing campaign aims, audience, descriptions, and different related info). We are able to use this listed information later to rapidly discover related reference pictures when creating new advertising and marketing campaigns. Under is a simplified implementation, with the whole code out there in our GitHub repository:

# Initialize the OpenSearch shopper
oss_client = OpenSearch(
    hosts=[{'host': OSS_HOST, 'port': 443}],
    http_auth=AWSV4SignerAuth(boto3.Session().get_credentials(), area, 'aoss'),
    use_ssl=True,
    verify_certs=True,
    connection_class=RequestsHttpConnection
)
# Put together doc for indexing
doc = {
    "id": image_id,
    "node": metadata['node'],
    "goal": metadata['objective'],
    "image_s3_uri": s3_url,
    "image_description": description,
    "img_element_list": components,
    "embeddings": embedding_vector
}
# Index doc in OpenSearch
oss_response = oss_client.index(
    index=OSS_EMBEDDINGS_INDEX_NAME,
    physique=doc
)

Combine the search engine into the advertising and marketing campaigns picture generator

The picture era workflow combines marketing campaign necessities with insights from earlier reference pictures to create new advertising and marketing visuals. The method begins when customers provoke a brand new marketing campaign via the net UI. Customers present three key inputs: a textual content description of their desired marketing campaign, its goal, and its node. Utilizing these inputs, we carry out a vector similarity search in OpenSearch Serverless to establish essentially the most related reference pictures from our library. For these chosen pictures, we retrieve their descriptions (created earlier via Lambda perform DescribeImgFunction) and incorporate them into our immediate engineering course of. The ensuing enhanced immediate serves as the muse for producing new marketing campaign pictures that align with each: the consumer’s necessities and profitable reference examples. Let’s look at every step of this course of intimately.

Get picture suggestions

When a consumer defines a brand new marketing campaign description, the Lambda perform GetRecommendationsFunction transforms it right into a vector embedding utilizing the Amazon Titan Multimodal Embeddings mannequin. By remodeling the marketing campaign description into the identical vector house as our picture library, we are able to carry out exact similarity searches and establish reference pictures that intently align with the marketing campaign’s aims and visible necessities.

The Lambda perform configures the search parameters, together with the variety of outcomes to retrieve and the okay worth for the k-NN algorithm. In our pattern implementation, we set okay to 5, retrieving the highest 5 most comparable pictures. These parameters may be adjusted to steadiness outcome range and relevance.

To assist guarantee contextual relevance, we apply filters to match each the node (audience) and goal of the brand new marketing campaign. This method ensures that really useful pictures should not solely visually comparable but additionally aligned with the marketing campaign’s particular objectives and audience. We showcase a simplified implementation of our search question, with the whole code out there in our GitHub repository.

physique = {
    "measurement": okay,
    "_source": {"exclude": ["embeddings"]},
    "question":
        {
            "knn":
                {
                    "embeddings": {
                        "vector": embedding,
                        "okay": okay,
                    }
                }
        },
    "post_filter": {
        "bool": {
            "filter": [
                {"term": {"node": node}},
                {"term": {"objective": objective}}
            ]
        }
    }
}
res = oss_client.search(index=OSS_EMBEDDINGS_INDEX_NAME, physique=physique)

The perform processes the search outcomes, that are saved in Amazon DynamoDB to keep up a persistent document of campaign-image associations for environment friendly retrieval. Customers can entry these suggestions via the UI and choose which reference pictures to make use of for his or her new marketing campaign creation.

Enhancing the meta-prompting approach with reference pictures

The immediate era section builds upon our meta-prompting approach launched in our earlier weblog. Whereas sustaining the identical method with Amazon Nova Professional 1.0, we now improve the method by incorporating descriptions from user-selected reference pictures. These descriptions are built-in into the template immediate utilizing XML tags (), as proven within the following instance.

You're a graphics designer named Joe that focuses on creating visualizations aided by text-to-image basis fashions. Your colleagues come to you each time they need to craft environment friendly prompts for creating pictures with text-to-image basis fashions corresponding to Secure Difussion or Dall-E. 

You at all times reply to your colleagues requests with a really environment friendly immediate for creating nice visualizations utilizing text-to-image basis fashions.

These are some guidelines you'll observe when interacting together with your colleagues:

* Your colleagues will talk about their concepts utilizing both Spanish or English, so please be versatile.
* Your solutions will at all times be in English whatever the language your colleague used to speak.
* Your immediate must be at most 512 characters. You're inspired to make use of all of them.
* Don't give particulars about or decision of the photographs within the immediate you'll generate.
* You'll at all times say out loud what you're pondering
* You at all times purpose solely as soon as earlier than making a immediate
* It doesn't matter what you at all times present a immediate to your colleagues
* You'll create just one immediate
* If supplied with reference picture descriptions (might be in between  XML tags) fastidiously steadiness the contributions of the campaigns description with the reference pictures to create the immediate
* By no means counsel so as to add textual content to the photographs

Listed below are some tips you at all times observe when crafting efficient picture prompts:

* Begin with a transparent imaginative and prescient: Have a transparent concept of the picture you need the AI to generate, picturing the scene or idea in your thoughts intimately.
* Select your topic: Clearly state the primary topic of your picture, making certain it's prominently talked about within the immediate.
* Set the scene: Describe the setting or background, together with the atmosphere, time of day, or particular location.
* Specify lighting and environment: Use descriptive phrases for lighting and temper, like "bathed in golden hour mild" or "mystical environment".
* Incorporate particulars and textures: Enrich your immediate with descriptions of textures, colours, or particular objects so as to add depth.
* Use adverse key phrases correctly: Embrace particular components you need the AI to keep away from to refine the output.
* Be aware of size and readability: Efficient prompts are typically detailed however not overly lengthy, offering key visible options, kinds, feelings or different descriptive components.
* Particular tokens may be added to offer higher-level steerage like "photorealistic", "cinematic lighting" and so forth. These act like key phrases for the mannequin.
* Logically order immediate components and use punctuation to point relationships. For instance, use commas to separate unbiased clauses or colons to guide into an outline.
* Evaluate and revise: Verify your immediate for accuracy and readability, revising as wanted to raised seize your concept.

Listed below are some examples of prompts you've created beforehand to assist your colleagues:

{Textual content to picture immediate examples}

A colleague of yours has come to you for assist in making a immediate for:

{textual content}

He additionally discovered the next picture descriptions that match what he wish to create and he desires you to contemplate the for crafting your immediate:


{Descriptions of associated reference pictures}

Utilizing your data in text-to-image basis fashions craft a immediate to generate a picture on your colleague. You're inspired to suppose out loud in your inventive course of however please write it down in a scratchpad.

Construction your output in a JSON object with the next construction:

{json_schema}

The immediate era is orchestrated by the Lambda perform GeneratePromptFunction. The perform receives the marketing campaign ID and the URLs of chosen reference pictures, retrieves their descriptions from DynamoDB, and makes use of Amazon Nova Professional 1.0 to create an optimized immediate from the earlier template. This immediate is used within the subsequent picture era section. The code implementation of the Lambda perform is out there in our GitHub repository.

Picture era

After acquiring reference pictures and producing an enhanced immediate, we use the Lambda perform GenerateNewImagesFunction to create the brand new marketing campaign picture. This perform makes use of Amazon Nova Canvas 1.0 to generate a last visible asset that comes with insights from profitable reference campaigns. The implementation follows the picture era course of we detailed in our earlier weblog. For the whole Lambda perform code, see our GitHub repository.

Creating a brand new advertising and marketing marketing campaign: An end-to-end instance

We developed an intuitive interface that guides customers via the marketing campaign creation course of. The interface handles the complexity of AI-powered picture era, solely requiring customers to offer their marketing campaign description and primary particulars. We stroll via the steps to create a advertising and marketing marketing campaign utilizing our answer:

1. Customers start by defining three key marketing campaign components:
   1. Marketing campaign description: An in depth transient that serves as the muse for picture era.
   2. Marketing campaign goal: The advertising and marketing intention (for instance, Consciousness) that guides the visible technique.
   3. Goal node: The particular viewers phase (for instance, Clients) for content material focusing on.

2. Based mostly on the marketing campaign particulars, the system presents related pictures from earlier profitable campaigns. Customers can evaluate and choose the photographs that align with their imaginative and prescient. These alternatives will information the picture era course of.

3. Utilizing the marketing campaign description and chosen reference pictures, the system generates an enhanced immediate that serves because the enter for the ultimate picture era step.

4. Within the last step, our system generates visible property based mostly on the immediate that would probably be used as inspiration for an entire marketing campaign briefing.

How Bancolombia is utilizing Amazon Nova to streamline their advertising and marketing marketing campaign property era

Bancolombia, one in every of Colombia’s main banks, has been experimenting with this advertising and marketing content material creation method for greater than a 12 months. Their implementation supplies useful insights into how this answer may be built-in into established advertising and marketing workflows. Bancolombia has been capable of streamline their inventive workflow whereas making certain that the generated visuals align with the marketing campaign’s strategic intent. Juan Pablo Duque, Advertising Scientist Lead at Bancolombia, shares his perspective on the impression of this know-how:

“For the Bancolombia crew, leveraging historic imagery was a cornerstone in constructing this answer. Our aim was to immediately deal with three main trade ache factors:

• Lengthy and expensive iterative processes: By implementing meta-prompting methods and making certain strict model tips, we’ve considerably diminished the time customers spend producing high-quality pictures.
• Problem sustaining context throughout inventive variations: By figuring out and locking in key visible components, we guarantee seamless consistency throughout all graphic property.
• Lack of management over outputs: The suite of methods built-in into our answer supplies customers with a lot better precision and management over the outcomes.

And that is only the start. This train permits us to validate new AI creations towards our present library, making certain we don’t over-rely on the identical visuals and maintaining our model’s look contemporary and interesting.”

Clear up

To keep away from incurring future prices, you must delete all of the assets used on this answer. As a result of the answer was deployed utilizing a number of AWS CDK stacks, you must delete them within the reverse order of deployment to correctly take away all assets. Observe these steps to wash up your atmosphere:

1. Delete the frontend stack:

2. Delete the picture era backend stack:

cd ../backend-img-generation
cdk destroy

3. Delete the picture indexing backend stack:

cd ../backend-img-indexing
cdk destroy

4. Delete the OpenSearch roles stack:

cd ../create-opensearch-roles
cdk destroy

The cdk destroy command will take away most assets robotically, however there could be some assets that require handbook deletion corresponding to S3 buckets with content material and OpenSearch collections. Be certain that to examine the AWS Administration Console to confirm that each one assets have been correctly eliminated. For extra details about the cdk destroy command, see the AWS CDK Command Line Reference.

Conclusion

This put up has introduced an answer that enhances advertising and marketing content material creation by combining generative AI with insights from historic campaigns. Utilizing Amazon OpenSearch Serverless and Amazon Bedrock, we constructed a system that effectively searches and makes use of reference pictures from earlier advertising and marketing campaigns. The system filters these pictures based mostly on marketing campaign aims and goal audiences, serving to to make sure strategic alignment. These references then feed into our immediate engineering course of. Utilizing Amazon Nova Professional, we generate a immediate that mixes new marketing campaign necessities with insights from profitable previous campaigns, offering model consistency within the last picture era.

This implementation represents an preliminary step in utilizing generative AI for advertising and marketing. The whole answer, together with detailed implementations of the Lambda capabilities and configuration information, is out there in our GitHub repository for adaptation to particular organizational wants.

For extra info, see the next associated assets:

Concerning the authors

María Fernanda Cortés is a Senior Information Scientist on the Skilled Providers crew of AWS. She’s centered on designing and creating end-to-end AI/ML options to handle enterprise challenges for purchasers globally. She’s captivated with scientific data sharing and volunteering in technical communities.

David Laredo is a Senior Utilized Scientist at Amazon, the place he helps innovate on behalf of shoppers via the appliance of state-of-the-art methods in ML. With over 10 years of AI/ML expertise David is a regional technical chief for LATAM who continually produces content material within the type of blogposts, code samples and public talking classes. He presently leads the AI/ML professional group in LATAM.

Adriana Dorado is a Pc Engineer and Machine Studying Technical Area Neighborhood (TFC) member at AWS, the place she has been for five years. She’s centered on serving to small and medium-sized companies and monetary providers clients to architect on the cloud and leverage AWS providers to derive enterprise worth. Outdoors of labor she’s captivated with serving because the Vice President of the Society of Ladies Engineers (SWE) Colombia chapter, studying science fiction and fantasy novels, and being the proud aunt of a wonderful niece.

Yunuen Piña is a Options Architect at AWS, specializing in serving to small and medium-sized companies throughout Mexico to rework their concepts into modern cloud options that drive enterprise progress.

Juan Pablo Duque is a Advertising Science Lead at Bancolombia, the place he merges science and advertising and marketing to drive effectivity and effectiveness. He transforms advanced analytics into compelling narratives. Enthusiastic about GenAI in MarTech, he writes informative weblog posts. He leads information scientists devoted to reshaping the advertising and marketing panorama and defining new methods to measure.