N@TM Survey Feedback: Bathroom Pass Scanner
To refine our system, we gathered feedback from our showcase visitors. Below, we have collected all direct feedback items received specifically for the Bathroom Pass & Biometric Scanner pipeline.
Visitor Feedback Table
| Feedback Received | Category | Plan & Actionable Response |
|---|---|---|
| It could be changed to be more interactive and user friendly | UI/UX Interaction | Polish the scanner interface with dynamic instructions and clear status displays. |
| This one has a lot of potential like the design is sorta basic and there could be more features | Design & Features | Rebuild UI with modern Glassmorphic styling and integrate richer visual elements. |
| Reduce ethnic bias | Biometrics | Critical: Normalizing illumination and dynamic cosine thresholds to ensure equitable facial matches across skin tones. |
| Add a button for an instruction panel in the corner | UI/UX Help | Add a persistent instruction panel/card in the screen corner explaining how to scan or enroll. |
| Make the blink to click feature for head tracker more precise and have a safety feature for not having too many people on the queue. | Biometrics & Queue Safety | Critical: Eye Aspect Ratio (EAR) calibration over 4 frames to ensure blink precision, combined with a hard limit (e.g., 5 students) on the FIFO queue. |
| Looks good | Appreciation | Positive visitor reinforcement. |
| You can add a feature to count number of absence for students | Classroom Analytics | Integrate student absence counts into the database model to help teachers monitor attendance trends. |
| Add trackers and other metrics | Classroom Analytics | Expand administrative metrics to capture weekly check-out averages. |
| One feature I would say that might be potential is the addition of a feature where if the person doesn’t sign out there is some form of ring that notify the person to sign out | Queue Alerts | Critical: Trigger a dashboard alarm/bell tone if a checked-out pass exceeds 10 minutes. |
| i really enjoyed this! make it more accessible. add voice commands. | Accessibility | Add HTML5 Web Speech API commands to enable voice-activated check-in fallbacks. |
| How will it identify multiple heads? | Biometrics | Configure scanner to lock matches strictly onto the largest detected face bounding box in frame. |
| I really like your bell ringing feature | Appreciation | Positive recognition of the automated end-of-period reset queue clearer. |
| Perhaps add features for scrolling up and down the form and use hand features on the head tracker. | Gesture Navigation | Integrate MediaPipe Hands to support hand-gestures (swipes/pinches) for touchless UI control. |
| Awesome | Appreciation | Positive visitor reinforcement. |
| Add a more user friendly ui | UI/UX Design | Transition from simple HTML layout components to a premium HSL styled CSS system. |
Architectural Iterations & Engineering Roadmap
Using the categorized feedback above, we have mapped out exact technical upgrades for the Bathroom Pass Scanner system. This demonstrates our commitment to secure, accessible, and highly refined biometric software.
1. Biometric Scanner Precision & Multi-Head Filtering
- The Feedback: “How will it identify multiple heads?” & “Make the blink to click feature more precise.”
- Our Technical Roadmap:
- Single-Subject Locking: We will configure the face-detection canvas to only run matches on the largest detected face bounding box in the frame (indicating the student directly in front of the screen). All secondary background faces will be ignored.
- Blink Click Calibration: We will refine the blink-trigger threshold by tracking the Eye Aspect Ratio (EAR) across 4 consecutive frames (approx 120ms duration) instead of instant single-frame drops. This distinguishes normal blinking from intentional triggers and avoids accidental check-ins.
2. Reminders, Alerts & Stale-Pass Prevention
- The Feedback: “If the person doesn’t sign out there is some form of ring that notify the person to sign out.”
- Our Technical Roadmap:
- Active Session Warning (The Ringback): While our Bell Ringer automatically flushes the queue at period boundaries, we will implement an active session timer. If a student pass is checked out for more than 10 minutes, the dashboard will flash red and play a gentle audio alert (a bell chime) to remind the teacher/student to clear the session.
3. Ethical AI: De-Biasing the Biometric Pipeline
- The Feedback: “Reduce ethnic bias.”
- Our Technical Roadmap:
- Adaptive Image Preprocessing: To eliminate matching failures due to lighting differences across skin tones, we will integrate CLAHE (Contrast Limited Adaptive Histogram Equalization) to normalize brightness and contrast before the face vector is sent to the DeepFace model.
- Optimized Distance Thresholds: We will transition from a static Euclidean distance threshold to a dynamic Cosine similarity check, which provides significantly more robust and equitable classification under varying lighting environments.
4. Advanced Gesture Control & Gesture Navigation
- The Feedback: “Perhaps add features for scrolling up and down… and use hand features on the head tracker.”
- Our Technical Roadmap:
- Gesture Mapping: We plan to integrate the MediaPipe Hands library on the client scanner page. This will let students navigate the UI touch-free by scrolling through class lists using a hand-swipe gesture and selecting options by holding a pinch for 1 second.
5. Accessibility & Interface Polish
- The Feedback: “Make it more accessible. Add voice commands.” & “Add a button for an instruction panel.”
- Our Technical Roadmap:
- Voice Integration: We will integrate the HTML5 Web Speech API to allow students to say “Scan check-in” or “Clear pass” as an optional accessibility fallback.
- Premium Glassmorphism Design: We will rebuild the basic CSS layout using structured SCSS styling. We will add a floating instruction panel card in the top-right corner with smooth hover micro-animations, a clean dark/light HSL toggle, and premium glassmorphic visual states to make the scanner feel cutting-edge.