Intelligent Solution for Restricted Area Recognition - CDL Exam Answer

Proposing an Effective, Cost-Efficient, and Low Maintenance Overhead Solution

Question

You are working on proposing a solution to a prospective security service provider that wants to implement an Intelligent solution in restricted areas of bank (like locker etc.) for recognizing people who are not authorized to access the area. Since the contract will be awarded to the google cloud compatible proposal that is most effective, cost-efficient, and has minimum maintenance overhead.

Which of the below would you propose?

Answers

Explanations

Click on the arrows to vote for the correct answer

A. B. C. D.

Correct Answer: C.

Option A is Incorrect.

OpenCV could be used to write code for the given purpose, but it is an open-source library.

Hence it will be less preferred over Cloud Vision AI.

Option B is Incorrect.

Rekognition is not a Google Cloud offering but is offered by AWS.

Option C is Correct.

Face detection using Cloud Vision AI on Google cloud is:

-Effective,

-Cost-efficient and has.

-Minimum maintenance overhead.

Option D is Incorrect.

This could be implemented, but it requires high expertise and higher maintenance and hence is not preferred.

https://opencv.org https://cloud.google.com/vision

In this scenario, a security service provider wants to implement an intelligent solution in restricted areas of a bank to recognize unauthorized individuals trying to access the area. The solution needs to be effective, cost-efficient, and have minimum maintenance overhead.

The following are the proposed solutions:

A. Propose to use OpenCV:

OpenCV (Open Source Computer Vision Library) is an open-source computer vision and machine learning software library. It can be used to develop real-time computer vision applications. OpenCV can detect and recognize faces in images and videos, and it can also be used for object detection, tracking, and segmentation. However, it requires significant programming knowledge to implement and maintain the solution.

B. Propose to implement Rekognition:

Rekognition is an AI-based image and video analysis service offered by Amazon Web Services (AWS). It can detect faces, objects, and scenes in images and videos. It can also analyze images and videos for inappropriate content. Rekognition is a reliable and widely used service but has a significant cost associated with it.

C. Propose to implement Cloud Vision AI:

Cloud Vision AI is an image analysis service offered by Google Cloud. It can detect and recognize faces in images and videos. It can also identify objects, landmarks, and logos in images. Cloud Vision AI has a powerful text recognition feature that can detect and extract text from images. It is easy to implement, and it has low maintenance overhead.

D. Propose to implement triplet loss function trained deep CNN model:

Triplet loss function trained deep CNN model is an advanced AI-based image analysis solution. It uses a deep convolutional neural network to detect and recognize faces in images and videos. The model is trained using triplet loss function to minimize the distance between anchor, positive, and negative faces. It has a high accuracy rate but requires significant resources for training and maintenance.

Considering the criteria of effectiveness, cost-efficiency, and minimum maintenance overhead, the best proposal would be to implement Cloud Vision AI. It is an image analysis service offered by Google Cloud, and it can detect and recognize faces in images and videos, identify objects and landmarks, and extract text from images. It is easy to implement and has low maintenance overhead, making it a cost-efficient and effective solution for the security service provider.