CSP Video Analytics & AI Monitoring 2 — Questions and Answers
Question 1: Which video analytics technique is most effective for detecting loitering in a defined zone?
- Object classification
- Dwell-time analysis (Correct answer)
- Edge detection
- Histogram equalization
Correct answer: Dwell-time analysis
Dwell-time analysis tracks how long an object remains within a defined region of interest and triggers alerts when the threshold is exceeded.
Question 2: In AI-based video surveillance, a high false-positive rate primarily indicates which issue?
- Insufficient storage bandwidth
- Poor model precision (Correct answer)
- High recall score
- Excessive frame rate
Correct answer: Poor model precision
A high false-positive rate reflects low precision, meaning the model incorrectly flags non-events as security incidents too often.
Question 3: What is the primary purpose of a 'virtual tripwire' in video analytics?
- Calibrating lens distortion
- Triggering alerts when objects cross a defined line (Correct answer)
- Compressing video streams
- Balancing camera exposure
Correct answer: Triggering alerts when objects cross a defined line
A virtual tripwire is a configurable line in the video frame that triggers an alert whenever an object crosses it in a specified direction.
Question 4: Which AI model architecture is most commonly used for real-time object detection in surveillance systems?
- Recurrent Neural Network (RNN)
- Generative Adversarial Network (GAN)
- Convolutional Neural Network (CNN) (Correct answer)
- Restricted Boltzmann Machine (RBM)
Correct answer: Convolutional Neural Network (CNN)
CNNs excel at image feature extraction and are the foundation of most real-time object detection frameworks like YOLO and SSD used in surveillance.
Question 5: When deploying edge-based video analytics, what is the primary advantage over cloud-only processing?
- Lower camera cost
- Reduced latency and bandwidth consumption (Correct answer)
- Higher video resolution
- Longer cable runs
Correct answer: Reduced latency and bandwidth consumption
Edge processing analyzes video locally on the device, reducing the need to transmit large video streams to the cloud and enabling near-real-time alerts.
Question 6: A surveillance AI system has 90% recall. What does this metric indicate?
- 10% of actual events are missed by the system (Correct answer)
- 90% of all detections are correct
- The system generates alerts 90% of the time
- 10% of frames are processed
Correct answer: 10% of actual events are missed by the system
Recall (sensitivity) of 90% means the system correctly detects 90% of actual events, but misses 10% of real incidents (false negatives).
Question 7: Which privacy-preserving technique allows video analytics to detect people without retaining identifiable facial images?
- Silhouette-based detection (Correct answer)
- 4K upscaling
- JPEG compression
- Infrared illumination
Correct answer: Silhouette-based detection
Silhouette-based or skeleton detection extracts body shape and movement without capturing facial features, protecting individual privacy.
CSP Video Analytics & AI Monitoring 2
Which video analytics technique is most effective for detecting loitering in a defined zone?