Success through vision

We are excited to introduce the Vision feature in nele.ai version 1.7.0. This feature enables advanced AI models from OpenAI and Microsoft Azure, such as GPT-4o and GPT-4 Turbo, as well as all available Claude 3 models, to generate text based on image content recognition. Curious? Let’s explore what this feature has in store for you.

‍

Availability and use of the Vision feature

The Vision feature is available with the new nele.ai version (1.7.0). You can upload image files in supported formats directly into your chat sessions with the aforementioned AI models. The models will then generate a description of the images upon request. This saves time and opens up a variety of new use cases to optimize your workflows.

‍

Important to know

Since Vision has only recently become publicly available, we want to keep expectations for the system realistic. For this reason, we have compiled a list of what users can and cannot expect from the current version of Vision:

What users can expect

Image description: When a user uploads an image, the system can provide a description of it. This is useful for understanding the content of an image or providing context for a discussion.

Object recognition: Vision can identify, name, and categorize specific objects within an image.

Scene recognition: The system can describe the general scene or context of an image, such as "a beach at sunset" or "a busy city street."

What users should not (yet) expect

Perfect accuracy: Although the system generally describes image content well, it is not entirely error-free. It may occasionally misidentify or misinterpret content or overlook seemingly unimportant details.

Personal identification: For data protection reasons, the system is unable to identify specific individuals in an image.

Integration into knowledge bases: Images cannot currently be uploaded to knowledge bases.

Reading images within documents or PDFs: By default, images can only be recognized as individual files. Images within documents cannot be recognized, as only text content is interpreted in documents.

Please note that Vision cannot display images or reference them. Alongside DALL-E, Vision is a separate application that operates independently of text models like ChatGPT or Claude 3. Vision can only provide descriptions and interpretations based on the information provided by the Vision system.

Use cases

Below are some inspiring use cases for Vision to spark your creativity and serve as a starting point. For more complex applications, integrating the nele.ai API into your own systems is required to enable batch processing and automated integration with other applications, thereby realizing extremely fast and productive workflows.

‍

Document management and archiving

Automatic image labeling: Automatically add descriptive text to images in your documents or archives to facilitate organization and searchability.

Image captioning: If you need an image description on the fly while creating a text document or presentation, you can generate one directly using the vision feature with a simple command.

‍

Prompt example

You are professional archiving software. Your goal is to create a short image description for the attached image so that it can be archived. Use the following format: [Suitable image name – Short description of the content shown] [06/17/2024]

‍

E-commerce and retail

Product description: Save valuable time by automatically generating product descriptions for your online shop based on product images.

Product categorization: Make navigation easier for your customers by having products automatically assigned to the appropriate categories.

Prompt example

You are a professional content manager for online shops. Your goal is to assign the content shown in the attached image to one of the following categories. Provide only the name of the appropriate category in your response, without any further notes or comments. Categories: Electronics, Fashion, Toys, Groceries

Image recognition in social media

Brand and logo tracking: Identify and describe brand logos in images shared on social media to analyze your brand's visibility.

Content moderation: Support your community moderation efforts by automatically detecting and describing image content.

‍

Prompt example

Prompt 1: You are a professional image analysis tool. Your goal is to identify a specific logo in images. Please memorize the attached logo for this purpose and confirm this to me.
Prompt 2: Now analyze the image attached here and tell me whether the logo from the previous step is shown in this image. Answer only with "Yes" or "No", without any further notes or comments.

‍

Marketing and advertising

Content analysis: Improve your advertising campaigns by analyzing visual content and increasing its effectiveness.

Personalization: Automatically create personalized marketing messages based on the analyzed image content.

‍

Prompt example

You are a professional marketing expert. Your goal is to create a personalized marketing text for the attached poster for the specified target group. The text should be short and concise so that it can be added to the poster. The target group to address is: Young adults and young professionals.

‍

Real estate and construction

Real estate listings: Create detailed descriptions for real estate photos in your listing portals and increase the appeal of your offers.

Construction progress: Document and describe construction progress with regular photos to gain a better overview of your project.

‍

Prompt example

Prompt 1: You are a professional building surveyor. Your task is to document the construction progress based on the attached photos. Use the following format: [Appropriate title for the current state of the building – Brief description of current construction progress compared to the previous version] [10.06.2024]
Prompt 2: Attached is a photo showing the current construction progress. Use today's date: [17.06.2024]

The vision feature of nele.ai is revolutionizing how we work with visual content. Are you ready to try out this powerful tool? Discover the efficiency and versatility of image content recognition and experience for yourself how it can optimize your processes. Try it out today!