Skip to main content
Using AI Visual Search

Find images by their visual content using AI

Piers Lowe avatar
Written by Piers Lowe
Updated over 2 weeks ago

OpenAsset’s AI Visual Search allows you to find images based on their visual content—no need for manual keywords. Search for terms like “Curving staircase,” “Forklift,” or “People on site.”

It can also surface results for more technical terms used in AEC, such as “Concrete pour,” “Beam,” or “Drywall.”

AI Visual Search is designed to work seamlessly with project-based keywords. Users can combine AEC-specific terminology like Industry Sector or Sustainability Rating with visual searches, ensuring that the results are not only visually accurate, but also relevant to the specific task at hand.

This tool significantly speeds up the upload process, as there is no need to spend time tagging what's visually in the image. This allows your team to focus on more creative and strategic tasks.

Quick Demo video:

Admins: Controlling access to Visual Search using Groups in OpenAsset:

Key Use Cases

1. Proposal and Presentation Creation:

Quickly locate relevant images like “high-rise at sunset” or “construction worker on-site” for client-facing documents.

2. Marketing Campaigns:

Discover visually compelling assets with queries like “prefabricated steel structure” or “modern glass building.”

3. Portfolio and Project Showcases:

Use AI Visual Search to curate recent, relevant images for portfolio updates and client presentations.

Why Still Build Your Taxonomy

Project keywords:

It is highly recommended to establish and maintain a project keyword taxonomy to capture important details such as location, sector, service, etc. This helps your team efficiently filter projects based on these criteria, and this data is often exported onto templates.

By combining visual search terms with project keywords, you can refine your searches and find exactly what you need. For example, you might search for “ribbon cutting” images (via visual search) specific to projects in the “Residential” sector located in “NYC” (using project keywords).

File keywords:

The AI Visual Search tool automatically searches your images based on their visual content, so you won’t need file keywords for obvious elements like:

  • Crane

  • Skyscraper

  • Furniture

  • Lighting

  • Foliage

However, file keywords are valuable for tagging elements that may not be visually recognizable in a photo. For example, specific engineering features, architectural styles, or other unique attributes that are important to your workflow can be tagged with file keywords to improve searchability.

Visual Search Permissions

AI Visual Search is made available to all users by default. However, you are able to control who has access if you wish. The feature can be controlled by permissions at the Group level.

To do this, follow these steps:

  • Go to Settings

  • System Settings

  • Users > Groups > [the group you would like to control]

  • Advanced > Turn “Use AI Visual Search” on or off.

  • Click “Save Changes.

If you would like to deactivate AI Visual Search for all users in your OpenAsset system, perform the steps listed above for each Group.

Visual Search Mode

Using visual search requires you to switch to visual search mode. You will notice a toggle in the search bar at the top of your OpenAsset site:

Switching to the “sparkle” AI icon will put you into visual search mode:

Now you can run a visual search. Type the words or phrase that describe the images that you are looking for and hit return:

If you don't find what you're looking for, try adding more context or changing the phrasing. Just like with other AI tools, it helps to be as descriptive as possible. It’s also worth scrolling down or paging through your search results to see if what you are looking for is not at the top of the screen.

Alternatively, you might find a photo you love and want to find more like it. Or you may find an image that's close to what you're looking for, but not quite it.

  • Click on the image

  • Click on the "Similar" tab (with the sparkly rainbow "AI" icon next to it)

  • Here you will see photos similar to the one you have selected.

Using Visual Search in combination with Regular Search

You can combine visual searches with regular searches in order to find exactly what you need.

For example, you might want to find “construction at sunset” images (visual search) but for projects in the "Healthcare" sector, in “Chicago” (regular search).

To do this:

Switch to visual search mode:

Type your visual search and hit enter:

Switch to regular search mode:

Add more search criteria in the usual way. For example, project keywords that describe the type of project:

Sort Order with Visual Search

When you have a visual search set, the search results that you see will be fixed at “Visual Relevance”. This is because relevance ordering is a fundamental part of the visual search functionality. The goal is for the most relevant results to be at the top of the screen.

You may have a scenario where you apply a visual search, and you then want to order those results by another criteria such as date created. To do this, make a Selection of the files that you want to order, then order the Selection.

How does the AI work?

OpenAsset’s visual search is powered by a third party AI model. The model is run within OpenAsset and within AWS so your data is not being sent to a third party. Your data remains within the secure boundary of OpenAsset.

The AI model is pre-trained on 400 million images. OpenAsset has not trained the model, nor is the model being trained as you use it. In that sense, we don't have a direct influence over the AI. That said, OpenAsset will be reviewing the performance of this feature and the value that it delivers to our clients to take opportunities to deliver improved models in the future.

In order for this feature to work, OpenAsset automatically “indexes” your images using the AI model. This index provides much richer information than traditional keywords. This means you can search with a much wider vocabulary while saving your company many hours of manual tagging.

Search Performance

Visual search performance can vary and is subjective, so exact results may not always be guaranteed. This variability can arise for several reasons:

  • Good search results will depend on the images in your system. You may not have what you are looking for.

  • The model will have a better understanding of some terms than others. It’s therefore worth trying other terms that might get you to the results that you are looking for.

  • Some searches may return relevant images, but the performance might not always be perfect. It’s a good idea to scroll or page through your results to explore further options.

  • The model has not been trained on AEC-specific terminology. However, it should perform well with commonly used terms related to buildings, engineering, and construction. For example, terms like “rendering,” “aerial shot,” “excavation,” “crane,” “bridge,” “tunnel,” “stairs,” and “kitchen” should work effectively. For highly specialized technical terms, the model may be less reliable, so we recommend using traditional file keywords for searches involving those terms.

  • Searches are likely to work better with more context, and tend to work better with phrases than with single words. Providing more context by describing what you’re looking for in a phrase will likely yield more accurate results compared to using a single-word search.

Did this answer your question?