New Body detection model and other Audience Analytics updates

Stage 1 of the Body Detection model

We’re proud to announce the launch of Stage 1 of our Body Detection model! This marks a huge step forward for our product in terms of functionality and accuracy. DISPL's Visitor Insights is now a unique solution running two AI models simultaneously: Face Detection AND Body Detection.

It’s just the beginning, but it opens the door to much deeper, smarter, and accurate analytics.

Before

Contact duration (attention time) was calculated based only on face attributes.

If a face was visible in front of the sensor, we assumed the person was present and counted that time toward the track length.

If the person turned their head away or left the frame, we stopped tracking — even if they were still physically there.

Now‍

Contact duration is now calculated based on both face and body detection.

Even if a person turns their head, as long as their body is visible, we will continue tracking.

This results in much more accurate measurement of how long someone was actually standing in front of the sensor.

What this means for you

You now get more accurate traffic and engagement data for your product displays or brand zones — giving you powerful insights to test specific areas of your store, optimize your layout, and fine-tune your offline marketing campaigns.

Fine tuning feature

Before

We previously offered a basic High Sensitivity Mode to help balance detection and visitor tracking:‍

• When it was enabled: Registered all tracks longer than 2 seconds, even with poor angle or movement.‍
• When it was disabled: Only registered tracks longer than 5 seconds with a good head angle.
  

The challenge? If you needed different types of data (e.g., full traffic vs. only engaged visitors), you had to change the setting and start data collection from scratch.

Now

With the new Fine Tuning feature, we collect all data automatically — and you can filter it anytime in your Dashboard based on your needs:

• Quality Priority: Focus on highly engaged visitors — those standing right in front of the sensor, with longer fixation time (minimum 4 seconds) and proper viewing angle.
  ‍Perfect for analyzing real engagement in specific product zones.‍
• Quantity Priority: Include by-passers and short glances — for a complete picture of overall traffic.
  ‍Ideal for understanding total footfall and store flow.

What this means for you

From day one of your POC, you now have the flexibility to dive deep into your data.

• Want to know how many people passed by? → Choose Quantity Priority.‍
• Want to focus only on real engagement? → Switch to Quality Priority.
  

You get full control and richer insights, without having to recollect or reconfigure your setup.

All these updates are just Stage 1 in a series of exciting product upgrades. The implementation of Body Detection takes our analytics to an entirely new level of precision — unlocking even more powerful ways to analyze audience behavior.

👉 Log in to your platform to explore the new features — and stay tuned for more updates coming soon!