Practice Test Geeks(CSP) Certified Surveillance Professional Practice Test
← All CSP Flashcard Decks

Video Analytics & AI Monitoring Flashcards

7 cards from real CSP practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.

Read the first 7 Video Analytics & AI Monitoring flashcards as text

  1. What is the primary challenge of using facial recognition in large crowded environments for surveillance?

    Answer: Increased false match rates due to occlusion and pose variation

    Crowded scenes introduce occlusion, varied angles, and lighting changes that significantly increase false match rates in facial recognition systems.

  2. A CSP professional is evaluating a video analytics system for a parking facility. Which feature is MOST critical for this environment?

    Answer: License plate recognition (LPR)

    LPR automates vehicle identification and access control, making it the most operationally critical analytics feature for parking facility management.

  3. What does 'mAP' (mean Average Precision) measure in the context of AI object detection for surveillance?

    Answer: Overall detection accuracy across multiple object classes and thresholds

    mAP averages the precision-recall curve across all object classes and IoU thresholds, providing a comprehensive detection performance benchmark.

  4. Which type of analytics is BEST suited for monitoring social distancing compliance in a workplace?

    Answer: Person-to-person distance estimation

    Person-to-person distance estimation uses camera calibration and object detection to measure gaps between individuals and alert when thresholds are violated.

  5. In AI surveillance, 'transfer learning' refers to which practice?

    Answer: Adapting a pre-trained model to a new surveillance task with minimal new data

    Transfer learning reuses knowledge encoded in a model trained on a large dataset (e.g., ImageNet) and fine-tunes it for a specific surveillance task with limited labeled data.

  6. A surveillance AI system triggers an alert every time a flag blows in the wind. Which analytics setting should be adjusted?

    Answer: Sensitivity threshold or minimum object size filter

    Adjusting the sensitivity threshold or setting a minimum detected object size filter prevents small, irrelevant movements like flags from triggering false alerts.

  7. What is the role of 'ground truth data' in validating a video analytics AI model?

    Answer: Serving as the verified, labeled dataset against which model predictions are compared

    Ground truth is human-verified labeled data used as the definitive reference to measure how accurately the AI model's predictions match reality.