In short: Logo Detection scans images for your brand logos, including altered or disguised ones, so infringements are caught even when the seller leaves your brand name out of the text.
Understanding Logo Detection
Logo Detection identifies brand logos (wordmarks and figurative marks) as posts and websites enter Zeal. It works with Auto-Moderation rules to flag potential counterfeit activity and trademark infringement.
It scans post images for unauthorised use of your logos, even when the post omits your brand name in the text. This makes it possible to monitor visually driven marketplaces and content-heavy platforms at scale.
1. Key capabilities
Logo Detection closes the gap left by text-only searches. It's trained on specific logo variations and brand features, so it catches altered, obscured or disguised logos.
Bypass text evasion: catch posts through visual logo use alone, when brand names are absent from the text.
Scan visual platforms: monitor marketplaces and platforms where image content drives posts.
Identify disguised assets: detect altered or disguised logos using AI trained on specific logo variations and brand traits.
Review efficiently: detections are highlighted in the image and post views for clear, actionable review.
Monitor at scale: keep watch on your brand with automated visual monitoring across large datasets.
Learn more: Auto-Moderation · AI Image Feature Detection
2. How it works
Knowing how the model is trained helps you understand when detections start appearing for your brand.
Logo Detection uses machine learning models to analyse incoming content, and is deployed in two phases:
v0 Logo Detection: an initial model trained on a selected set of logos. Logos are identified and combined with official product images, and Large Language Models (LLMs) help by automatically detecting and labelling logos in these images.
v1 Logo Detection: a refined model with higher accuracy. A brand logo guide is produced to define the final logo set, and the dataset (v0 data plus additional images) is manually labelled to train v1.
Note: From deployment onwards, v1 applies to all incoming content. The Logo Detection insight becomes available in Zeal once training and deployment are complete.
Learn more: IP Assets · Image Feature Configuration
3. Find posts with logo detections
Filter for logo detections to review or enforce the posts flagged by the model.
Use either of these filters:
Insights filter: select Logo detected.
Reasons filter: select Logo detected.
Learn more: Insight Filter · Reasons Filter
Quick reference
If you want to know… | Look at… |
What Logo Detection can catch | 1. Key capabilities |
How the model is trained and when it applies | 2. How it works |
Which posts have a logo detected | 3. Find posts with logo detections |
How to label logo detections automatically |