Azure AI Video Indexer
An Azure video analytics service that uses AI to extract actionable insights from stored videos.
96 questions
Issue with Video Indexer Identifying Different People as the Same
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(294.40000000000003, 71%, 38%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EEI%3C/text%3E%3C/svg%3E)
eMAM IT
0
Reputation points
We have noticed that Video Indexer is identifying different people as the same person.
As seen in the screenshot below, the faces marked do not belong to UnknownPerson105. This is a direct screenshot from the Video Indexer interface itself.
Could you please look into this issue and provide a possible solution? Looking forward to your response.
{count} votes
-
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(124.80000000000001, 17%, 22%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EPP%3C/text%3E%3C/svg%3E)
Pavankumar Purilla 4,025 Reputation points Microsoft External Staff2025-03-07T21:53:07.56+00:00 Hi eMAM IT,
Greetings & Welcome to the Microsoft Q&A forum! Thank you for sharing your query.It looks like Azure Video Indexer is misidentifying different people as the same individual in its facial recognition results. This could be due to model limitations, low-quality frames, or similarities in facial features under certain conditions. Here are some steps to improve accuracy:
- Use the Insights tab in Video Indexer to manually separate wrongly identified faces.
- Analyze the JSON insights and adjust the threshold to exclude low-confidence matches.
- Manually add more diverse and clear images of the correct individuals if you have access.
- Ensure the input video meets the recommended criteria (e.g., 30 FPS, proper lighting, and avoiding too many people per frame).
- Ensure the detected individuals appear in the frame with a clear, unobstructed view of their face and clothing.
For more information: Get observed people detection and matched faces insights
I hope this information helps.
Sign in to comment
Sign in to answer