Please enable JavaScript.
Coggle requires JavaScript to display documents.
Software Engineering Project Topics (Group 2), Software Engineering…
- Software Engineering Project Topics (Group 2)
- Proposal 1: RECALL (Agentic Personal Memory & Security Vault)
- Area: Agentic AI, Mobile/Desktop Computing, Zero-Knowledge Security
- SDGs: SDG 4 (Education), SDG 8 (Decent Work), SDG 9 (Innovation), SDG 16 (Strong Institutions)
- Session-Based Input Capture
- Text: Optical Character Recognition (OCR) for on-screen content
- Audio: Speech-to-Text (STT) for spoken dialogue and lectures
- AI Processing & Management
- Agent Pipeline: FastAPI backend, Supabase DB, chain of free LLMs
- Automatic Categorization: Deadlines, lessons, and credentials
- Memory Retention: Forgetting-curve schedule ("Still remember" & "Remind me")
- Zero-Knowledge Privacy Architecture
- End-to-End Encryption: Wrapped data keys
- Privacy Guarantee: No operator access, no key-recovery backdoor
- Proposal 2: LearnLite (Offline-First Learning App)
- Area: Mobile Computing, Human-Computer Interaction (HCI), EdTech
- SDGs: SDG 4 (Education), SDG 10 (Reduced Inequalities), SDG 1 (No Poverty)
- Architecture & Accessibility
- Progressive Web App (PWA) framework
- One-time download for complete offline usability
- Target Subjects: Grade 7-10 Math, Basic Programming
- On-Device Evaluation: Quizzes scored locally on client device
- Adaptive Retention: Spaced-repetition scheduler using Leitner boxes
- Synchronization & Analytics
- Opportunistic Cloud Sync when online
- Instructor Dashboard: Real-time mastery tracking and topic alerts
-
- Proposal 3: Freshwater Fish Classification
- Area: Computer Vision, Deep Learning, Environmental Monitoring
- SDGs: SDG 14 (Life Below Water), SDG 15 (Life on Land)
- Object Detection: Multi-fish localized detection via YOLO
- Feature Extraction & Classification: PyTorch & EfficientNet-B0
- Zero-Shot Support: OpenAI CLIP integration
- Ecological & Application Impact
- Target Species: Local native vs. invasive freshwater species
- Core Function: Early detection, automated flagging, and biodiversity monitoring
- Platform: Web application built with Streamlit/Flask
- Software Engineering Project Topics (Group 2)
- Proposal 1: RECALL (Agentic Personal Memory & Security Vault)
- Area: Agentic AI, Mobile/Desktop Computing, Zero-Knowledge Security
- SDGs: SDG 4 (Education), SDG 8 (Decent Work), SDG 9 (Innovation), SDG 16 (Strong Institutions)
- Session-Based Capture: On-screen OCR & Speech-to-Text audio
- Agentic AI: FastAPI backend, Supabase, chain of open LLMs
- Forgetting-Curve Scheduling: "Still remember" & "Remind me" prompts
- Security: Zero-knowledge encryption with wrapped data keys
- Proposal 2: LearnLite (Offline-First Learning App)
- Area: Mobile Computing, Human-Computer Interaction (HCI), EdTech
- SDGs: SDG 4 (Education), SDG 10 (Reduced Inequalities), SDG 1 (No Poverty)
- Architecture: Progressive Web App (PWA) with one-time download
- Learning Engine: On-device quiz scoring & Leitner box spaced-repetition
- Instructor Sync: Opportunistic cloud sync for teacher analytics dashboard
- Proposal 3: Freshwater Fish Classification for Invasive Species Detection
- Area: Computer Vision, Deep Learning, Environmental Monitoring
- SDGs: SDG 14 (Life Below Water), SDG 15 (Life on Land)
- Detection Pipeline: YOLO multi-fish detection & PyTorch/EfficientNet classification
- Zero-Shot Support: OpenAI CLIP integration for flexible species identification
- Ecosystem Goal: Flag invasive species to support biodiversity monitoring
- Proposal 4: Memora Local (On-Device Assistant for Dementia Care)
- Area: Mobile Computing, Human-Centered AI, Assistive Health Technologies
- SDGs: SDG 3 (Good Health and Well-Being), SDG 10 (Reduced Inequalities)
- Sensor Fusion: Infers patient context from light, steps, and phone unlocks
- Privacy & Reliability: 100% offline routine engine using on-device TTS & small LLMs
- Secondary Features: On-demand face recall ("Who is this?") and caregiver portal
- Proposal 5: AI Gym Exercise Form Correction & Rep Tracking System
- Area: Computer Vision, Pose Estimation, Fitness Technology
- SDGs: SDG 3 (Good Health and Well-Being), SDG 9 (Industry & Innovation)
- Pose Tracking: Real-time skeletal joint tracking via MediaPipe / OpenPose
- Form Analysis: Biomechanical angle calculation to detect risky movement forms
- Feedback & Counting: Automated repetition counting and real-time audio guidance
- Proposal 6: AI Crop Disease & Pest Detection System
- Area: Computer Vision, Precision Agriculture, Applied Machine Learning
- SDGs: SDG 1 (No Poverty), SDG 2 (Zero Hunger), SDG 12 (Responsible Consumption)
- Visual Diagnosis: Leaf image classification for disease and pest identification
- Treatment Guidance: Instant actionable mitigation strategies for local farmers
- Deployment: Mobile/web interface optimized for field usability