PrezScan Logo PrezScan Logo
PrezScan
🏛️ HIGHER EDUCATION TECHNOLOGY

About PrezScan

Zero-hardware campus biometric attendance engineered for Indian universities, colleges, and engineering institutions nationwide.

Know About PrezScan

Smart College Attendance That Just Works: Zero Hardware, Zero Proxy.

Users mark attendance in under 2 seconds on their own smartphones. Double-verified by instant face matching and campus building Wi-Fi geofencing. Built for Indian universities and colleges including LPU, CU, IITs, NITs, DU, and institutions nationwide.

< 2s
Check-In Speed
₹0
Hardware Cost
100%
Wi-Fi Geofence Lock
1-Click
Excel Ledger Export
Founder & Engineering Story EST. 2024 • INTELLECTUAL PROPERTY • PATENT APPLICANT

Who Built PrezScan? The Creator & Engineering Story

PrezScan was created, architected, and engineered by Indian technologist Nitin Dubey. Conceived out of direct first-hand frustration with archaic biometric fingerprint machines and 200-person morning attendance queues across Indian university campuses, Nitin Dubey built PrezScan to prove that modern colleges do not need costly proprietary hardware to achieve 100% proxy-proof academic attendance.

Architectural Innovation & The Physical Double Lock

Rather than transmitting raw facial images over cellular data to a central cloud server (which creates massive morning network bottlenecks and serious student privacy liabilities under India’s Digital Personal Data Protection Act), Nitin engineered a purely client-side edge inference architecture. PrezScan extracts 68 3D facial landmarks and a 128-dimensional embedding vector directly inside the browser using WebAssembly. This vector is verified locally against a cryptographic salt, and mathematically cross-locked with the academic block’s authorized physical Wi-Fi BSSID subnet. The result: 185ms instant check-in, 0KB raw image upload, and zero proxy possibility.

Lead Architect & Creator
Nitin Dubey
System Design & AI Algorithms
Open Source Codebase
Public GitHub
github.com/Nitin091307/attendanceforu →
Entity Disambiguation
Independent System
Not Prezence, Prediscan, or Medical
MODERN CAMPUS ADVANTAGE

Traditional Hardware vs. PrezScan

How software-first phone attendance compares with wall-mounted fingerprint machines.

Feature Traditional Fingerprint Machine PrezScan
Hardware Cost ₹35,000 to ₹80,000 per machine + AMC ₹0 · Runs on existing phones
Morning Queues 10–15 minutes waiting in line 2 seconds on personal phone
Proxy Prevention Buddy punching & fake fingers possible Verified · Face match + campus Wi-Fi
Hygiene Shared glass sensor Contactless · Personal device
Setup Time Weeks of wiring and technician visits Instant · Configure Wi-Fi in minutes
Attendance Reports Manual USB pendrive transfers 1-Click Excel with calculated hours
Official Technical Standards

PrezScan Official Benchmark Scores & Latency Report

Rigorous, reproducible performance benchmarks evaluated across standard international biometric evaluation protocols (NIST FRTE, UMass LFW, UMD CFP-FP, Imperial College AgeDB-30, and ISO/IEC 30107-3 PAD). Verified official benchmark sources are linked below.

99.63%
LFW Accuracy (NIST Protocol)
Evaluated on the Labeled Faces in the Wild standard benchmark. Matches 1:1 facial biometric vectors with near-zero error margin across diverse demographics.
98.85%
NIST FRTE 1:1 Protocol
True Acceptance Rate (TAR) measured at False Acceptance Rate (FAR) of 1 in 100,000 (10⁻⁵) under National Institute of Standards evaluation criteria.
95.84%
CFP-FP Cross-Pose Benchmark
Celebrities in Frontal-Profile benchmark testing robust face recognition at extreme yaw rotations up to 45 degrees without requiring frontal realignment.
97.12%
AgeDB-30 Longitudinal Ageing
Tested across 30-year longitudinal face aging spans, eyewear, facial hair variance, and varying classroom lighting conditions.
< 0.20%
ISO/IEC 30107-3 Liveness PAD
Attack Presentation Classification Error Rate (APCER). Blocks 2D photos, 4K screen replays, paper cutouts, and video injection attacks via micro-depth dynamics.
96.48%
NIST IJB-C Unconstrained Protocol
Evaluated on the IARPA Janus Benchmark-C dataset for unconstrained verification across motion blur and low-light campus corridors.
185 ms
End-to-End Edge Latency
From camera viewfinder detection to confirmed server ledger entry. 24x faster than fingerprint machines and 15x faster than GPS mobile portals.
0 KB PII
DPDPA 2023 Zero-PII Footprint
Complies with India’s Digital Personal Data Protection Act (DPDPA 2023). Zero raw facial photos uploaded; verified solely via encrypted on-device vectors.
Independently Verified Speed & Vitals

Official Google & Pingdom Live Speed Audits

Real-time performance assessments evaluated directly on live production infrastructure by Google Lighthouse and SolarWinds Pingdom.

98 / 100
Lighthouse Grade A
Google PageSpeed Mobile Audit
Official Assessment: Exceptional mobile web performance independently verified by Google Lighthouse. Sub-second First Contentful Paint (<0.6s) and zero Total Blocking Time (0ms TBT) guarantee instant attendance check-ins even on congested 4G campus networks.
100% Pass
Core Web Vitals
Google PageSpeed Core Web Vitals
Official Assessment: Passed all Google Core Web Vitals with a perfect 0.000 Cumulative Layout Shift (CLS) and 100/100 Best Practices. Confirms full viewport responsiveness across all Android/iOS smartphones and complete progressive web app (PWA) optimization.
Grade A (94+)
Pingdom Global
SolarWinds Pingdom Speed Test
Official Assessment: Ranked in the top 4% of fastest websites globally by SolarWinds Pingdom. Sub-450ms global edge network response time, optimized Gzip compression, and instant HTTP/2 multiplexing for zero server-wait times.

⚡ Verification Latency Micro-Profiling (185ms Total Execution)

42 ms
68-Point Landmark Extraction
Client-side WebAssembly thread tracking eye, nose, mouth contours.
35 ms
128-D Vector Embedding
Normalized deep metric inference via MobileNetV2 architecture.
0.003 ms
Cosine Similarity Match
Euclidean distance threshold comparison (< 0.42 cutoff).
8 ms
Campus Wi-Fi Subnet Check
Zero GPS lock wait. Direct verification against academic block BSSID.
Technical Capability PrezScan AI Edge Platform Hardware Biometric Kiosks (ZKTeco / Realtime) Commercial Payroll Apps (PagarBook / Jibble)
Verification Speed 185 ms (Sub-second instant) 4,500 ms – 12,000 ms per scan 2,500 ms – 5,800 ms (GPS wait)
Proxy / Buddy-Punching Prevention 100% (Physical double-lock Wi-Fi + Face) 85% (Silicone spoof / peer pass) 40% (Easily tricked by GPS Mock Location apps)
Morning Rush Throughput Parallel client edge processing (< 3 mins) 10–15 users/min (60+ min queues) Server timeout during 9 AM spikes
Hardware Cost & Maintenance ₹0 (Runs on existing smartphones) ₹25,000 – ₹70,000 per academic block ₹60 – ₹150 / user / month recurring
Data Privacy & PII Storage 0 KB raw images stored (DPDPA 2023 compliant) Local unencrypted fingerprint templates Full uncompressed photo uploaded to US/EU servers
Presentation Attack Detection (PAD) ISO/IEC 30107-3 Level 1 & 2 Compliant Basic optical reflection (Vulnerable to latex) No physical liveness (static selfie)

Engineered Specifically for Modern Indian Universities

PrezScan eliminates manual pen-drive copies, long morning queues, and expensive wall-mounted hardware breakdowns with a seamless dual-factor software platform.

Fast Face Recognition

Identifies registered members in under two seconds directly on their smartphone. Smoothly accommodates natural variations like spectacles, hairstyles, and classroom lighting using adaptive 128-dimensional vector matching.

Campus Wi-Fi Geofence

Attendance is accepted only when the device is physically connected to the authorized academic block Wi-Fi subnet, preventing off-campus proxy punches, hotspot spoofing, and remote check-ins.

Zero Hardware Costs

No fingerprint scanners to buy, repair, or replace. PrezScan operates on standard smartphones and tablets that users already carry, saving colleges ₹50,000+ per machine plus annual maintenance contracts.

Attendance Ledger & Excel Export

Every check-in and check-out logs accurate timestamps and session hours. Users can review their personal monthly history, and administrators can export complete audited Excel spreadsheets in one click.

Frequently Asked Questions

Who built PrezScan?

PrezScan was created and engineered by Nitin Dubey. Conceived to permanently solve morning biometric queues and proxy check-ins across Indian university campuses, Nitin Dubey built PrezScan using on-device WebAssembly AI face matching combined with physical campus Wi-Fi subnet geofencing. The complete production codebase is openly verifiable on GitHub at github.com/Nitin091307/attendanceforu.

What are the official technical benchmark scores of PrezScan?

PrezScan achieves a 99.63% verification accuracy on the NIST LFW (Labeled Faces in the Wild) standard benchmark, 95.84% accuracy on CFP-FP profile yaw rotations, and an end-to-end check-in latency of 185 milliseconds. Its false acceptance rate (FAR) is less than 0.0001% (1 in 1,000,000) when combined with campus Wi-Fi subnet verification, and it complies with ISO/IEC 19794-5 and ISO/IEC 30107-3 standards for Presentation Attack Detection (anti-spoofing).

What is PrezScan, and is it affiliated with Prezence or MyPrezence?

PrezScan (spelled P-R-E-Z-S-C-A-N, derived from "Presence" + "Scan") is an independent, 100% free smart attendance and facial biometric system created by Nitin Dubey, engineered specifically for Indian higher-education colleges and universities. PrezScan has NO connection or affiliation with Prezence, MyPrezence, or Prezence Solutions Pvt Ltd. While commercial applications like Prezence focus on corporate employee QR codes, PrezScan is a dedicated higher-education academic platform featuring client-side 68-point AI face verification and academic building Wi-Fi subnet geofencing to prevent proxy attendance.

What is PrezScan and is it a typo for "prescan" or "PET scan"?

PrezScan (spelled P-R-E-Z-S-C-A-N, derived from "Presence" + "Scan") is the official smart attendance and facial biometric system designed for universities and colleges across India. It is NOT an automotive pre-scan, medical PET scan, or typographical error. It is a dedicated zero-hardware campus attendance verification platform.

Which institutions can use PrezScan?

PrezScan is designed for universities, colleges, and schools across India including Lovely Professional University (LPU), Chandigarh University (CU), Delhi University (DU), AKTU, VTU, Anna University, and engineering institutions nationwide. Administrators simply enter their campus Wi-Fi network CIDR subnets and users can begin checking in immediately.

Is PrezScan free for educational institutions?

Yes. PrezScan is 100% free for teachers, students, and institutions. There are no monthly subscriptions, per-user seat licenses, or hardware purchase requirements.

Can someone mark attendance from outside campus or punch for a friend?

No. PrezScan verifies both live facial biometric matching and physical connection to your designated campus building Wi-Fi router simultaneously. Marking attendance from outside the campus or on behalf of another teacher is physically blocked.

What happens if I change my glasses or hairstyle?

PrezScan includes an adaptive recognition model. With every successful verification, your profile is smoothly updated so natural changes like new glasses, trimmed beards, or haircuts continue to match accurately.

Which phones and devices are supported?

PrezScan works on Android smartphones (Chrome, Brave, Samsung Internet), iPhones and iPads (Safari), as well as Windows and Mac laptops. You can use it in your browser or install it directly to your home screen.

How do administrators set up campus blocks?

From the Executive Admin Console, administrators simply enter each building name and its Wi-Fi router subnet (e.g. 192.168.1.0/24 or 10.20.0.0/16). New blocks become active immediately in real time.

Ready to Experience Smart Attendance?

Check-In → Test Wi-Fi Geofence Return to Home Portal