Real-Time Object Detection TensorFlow Model

I am seeking a skilled professional who can take an existing caffe model and convert it into an efficient object detection TensorFlow model with the following requirements:

- Implement the model in Python: As my preferred programming language is Python, the model should be implemented in this language.
- Surveillance Application: The primary application of the model is surveillance. This means that the model should be designed to be used in a surveillance context.
- Real-Time Processing: The model must process in real-time. This means that it should be able to detect objects with minimal delay, crucial for surveillance applications.

Other requirements include:
- mAP, F1 Score & FPS Analysis: Provide a comprehensive evaluation of the model's performance through mAP, F1 score, and frames per second (FPS) analysis. This will help us gauge the model's accuracy and speed, two critical factors in its effectiveness.
- Model Explanation: Offer an in-depth explanation of the model architecture. I would like to understand the internal workings of the model and how it detects objects in real-time.

Ideal candidates for this project should have:
- Strong TensorFlow expertise: The professional should be well-versed in TensorFlow and have experience in converting models from other frameworks to TensorFlow.
- Object Detection Experience: A history of working on object detection models is crucial. Experience in surveillance applications would be a plus.
- Strong Python Skills: Given that the model should be implemented in Python, the professional should have excellent Python skills.
- Real-Time Processing Understanding: A good grasp of real-time processing requirements and implementations for machine learning models.