# Group 10

**Objective:** Develop a robot that can find and spy on specific documents while also functioning as a vacuum cleaner. The robot should start spying when no one is in the room or when no one is looking at it, and start vacuuming when it detects people.

#### Deep Learning:

1. **Presence Detection:**
   * **Task:** Detect whether people are in the room or looking at the robot.
   * **Technique:** Use deep learning models to identify human presence and their gaze direction.
2. **Document Identification:**
   * **Task:** Recognize and locate specific documents.
   * **Implementation:** Train a model to identify target documents based on visual features and keywords.

#### Robotics and Hardware:

1. **Vacuum Cleaning Functionality:**
   * **Task:** Develop a vacuum cleaning mechanism that can operate efficiently.
   * **Implementation:** Equip the robot with a vacuum cleaner that can be activated based on the presence detection model.
2. **Object Manipulation:**
   * **Task:** Design a mechanism for grabbing or vacuuming up items, including documents.
   * **Implementation:** Ensure the robot can securely grab or suction documents and other small objects.

#### Implementation Steps:

1. **Deep Learning Development:**
   * **Presence Detection:** Collect and annotate data to train models for detecting human presence and gaze direction.
   * **Document Identification:** Create a dataset of target documents and train the model to recognize and locate them.
2. **Robotics Development:**
   * **Vacuum Mechanism:** Design and build a vacuum cleaning system for the robot.
   * **Object Manipulation:** Develop a mechanism for grabbing or suctioning items.
3. **System Integration:**
   * **Presence-Based Behavior:** Integrate the presence detection model to switch between spying and vacuuming modes.
   * **Document Retrieval:** Implement the document identification model to guide the robot in locating and retrieving target documents.

#### Next Steps for Students:

1. **Deep Learning Tasks:**
   * **Dataset Preparation:** Collect and label data for presence detection and document identification.
   * **Model Training:** Train and validate models for each task, focusing on accuracy and reliability.
2. **Robotics Tasks:**
   * **Vacuum Design:** Design and prototype the vacuum cleaning system.
   * **Grabbing Mechanism:** Develop and test the mechanism for grabbing or suctioning documents.
3. **Testing and Iteration:**
   * **System Testing:** Conduct tests to ensure the robot can switch between spying and vacuuming modes based on presence detection.
   * **Iterative Improvement:** Refine the design and functionality based on test results.

#### Final Considerations:

* **System Integration:** Ensure smooth integration of deep learning models with the robotic hardware.
* **Reliability:** Focus on making the system reliable in detecting human presence and switching modes accurately.
* **User Safety:** Design the system to operate safely around humans and avoid any potential hazards.

#### Summary:

* Develop a robot that can spy on documents and vacuum the room based on the presence of people.
* Use deep learning for human presence detection and document identification.
* Design and integrate a vacuum cleaning system and a mechanism for grabbing or suctioning items.
* Ensure seamless switching between spying and vacuuming modes for effective operation.