Introduction

Automatic person detection is a crucial technology that can enhance surveillance systems' capabilities, particularly in high-crime areas. VIDIZMO's AI Indexer is an innovative solution that can automatically detect individuals in surveillance videos, improving the searchability of individuals within a video, even if they appear at different times and angles.


VIDIZMO's AI Indexer uses machine learning algorithms to detect individuals across a single class - "person" - in surveillance videos that may contain cross-crime scenes from CCTV footage. The technology allows the indexer to quickly and accurately read individuals within videos without the need for human intervention.


When a video is inputted into the indexer, it starts a workflow activity that takes some time to detect individuals in the video. The processing time may vary depending on the video quality, duration, and resolution. Once the detection is complete, the user can navigate to Studio Space and view all the detected individuals. The user can also redact any sensitive information using the redaction feature.


The ability to automatically detect individuals in surveillance videos can significantly improve surveillance systems' effectiveness. By automating the detection process, surveillance professionals can focus on more critical tasks, such as analyzing suspicious activities and identifying potential threats. With VIDIZMO's AI Indexer, organizations can enhance their surveillance capabilities and ensure a safer and more secure environment.


Concept

Automatic person detection employs machine learning algorithms to detect humans across a single class i.e., "person" in videos that frequently include cross crime scenes from CCTV footages mainly. This automatic Person Detection technique enables VIDIZMO indexer to read individuals quickly and automatically, without the need for human intervention.


VIDIZMO takes videos captured that may contain individuals or CCTV video recordings containing individuals as input. The indexer takes the video and detects individuals that appear within the video at different timestamps with different angle views and classifies them as "person".


VIDIZMO indexer takes video as input. Indexer starts a workflow activity that takes some time to detect individuals in video. Video processing time may vary as it depends on factors such as video quality, duration, and resolution. Once the detection is complete, the user can see the detected individuals by navigating to Studio Space and redact all the detected individuals.


To understand redaction in detail, please refer to this article: Understanding Redaction using Studio  Space. In order to redact the automatically detected individuals, please refer to this article: How to Redact PII from Digital Evidences.



Person Detection Improvement by 25%

We are thrilled to announce that our person detection AI model has been updated, resulting in a 25% increase in its accuracy.


This update means that our AI model is now even better at detecting people in images and videos, with fewer false positives and false negatives. This improvement is especially crucial for security and surveillance applications where accurate identification and tracking of people are essential.


Our commitment to delivering the best possible experience for our users drives us to constantly improve our product. With this update, you can expect more reliable and accurate identification of individuals in different types of videos, including CCTV, dashcam, body cam, and mobile captured footage.

  1. Improved accuracy: With the improved person detection, you can expect more accurate and reliable identification of individuals in different types of videos, such as CCTV, dashcam, body cam, and mobile captured footage.
  2. Better organization and management of digital evidence: The improved person detection can enable the efficient and effective organization and management of digital evidence, allowing you to more easily identify, classify, and access the relevant information.
  3. Enhanced collaboration: Improved person detection can help to facilitate collaboration and communication among different departments, organizations, and stakeholders, as well as enable more accurate and reliable information sharing and analysis.
  4. Better support for investigations: With improved person detection, you can more effectively extract information from different types of video footage, such as facial recognition, movement patterns, and demographics, which can be valuable in various investigations, such as criminal cases, accidents, and incidents.
  5. Increased efficiency: Improved person detection can result in more efficient and effective use of your VIDIMZO Digital Evidence Management System, reducing the time and resources required for manual review and analysis.

 

Tracker Performance Improvement

We're excited to announce that we've updated our person tracker to use the ByteTrack algorithm, resulting in a significant improvement in performance.

With this new update, our person tracker is now even more accurate and efficient at tracking people in video footage. By using the ByteTrack algorithm, we've been able to improve the precision and speed of our person tracker, which means that you can rely on our product to accurately track people in real-time.

This update also means that our product is now more optimized and uses fewer resources, making it a more lightweight and efficient solution for your tracking needs. Whether you're using our product for security, surveillance, or any other application, you'll notice a significant improvement in the speed and accuracy of our person tracker.


Please refer to the following article to understand what model size, confidence threshold, and tracking frames are: How to configure VIDIZMO Indexer for object detection. As always, if you have any questions or feedback, please don't hesitate to reach out to our support team.