NBOX Forensics
Video investigation and face search with natural language.
One device. No live cameras needed. Upload video files from any source - multiple feeds at once. Run face recognition, object detection, and natural language queries across all of it.
Matches faces against photographs up to 30 years old - including low-resolution voter ID quality images. Proven in Bihar elections.
Request BriefingKey Specs
Best for: Police stations, campus security, corporate investigations, compliance audits.
Natural Language Query Examples
"Find all red jackets near the exit between 9pm and midnight."
"Show every instance this face appears across all feeds."
"Find all vehicles near camera 3 after 10pm."
"List all people who entered through the north gate on Tuesday."
"Find all instances of two or more people standing together for more than 5 minutes."
"Show all faces that appear in more than three different camera feeds."
Frequently Asked Questions
Can face recognition match old or low-quality photographs?
Yes. FaceTagr Forensics matches faces against photographs up to 30 years old, including low-resolution voter ID quality images. Deployed in Bihar state elections for this exact purpose.
How does forensic video face recognition work?
Upload video footage from any source. NBOX Forensics processes files offline, runs face recognition and object detection, and supports natural language search queries across all footage.
What is NLP video search?
You describe what you are looking for in plain English - for example, 'find all people in a red jacket after 8pm' - and the system returns matching results across all camera feeds.