-
Infrastructure analyzed with object detection models and trained CNNs in real - time to provide information on upkeep of institute’s infrastructure quality
-
Query based tabular data extraction models + transformers for analyzing authenticated documents to verify compliance with standard requirements/norms
-
Integration with institutional database to cross-check data with existing records
-
ML algorithms implemented to analyze trends in current + historical data, to address potential flags
-
User-friendly dashboard with customizable inspection criteria to provide flexibility
-
Automatically generated comprehensive reports include compliance status, faculty analysis, improvement suggestions for any deficiencies found and more
- Image Recognition for Facility Analysis: YOLOv8 used to assess infrastructure by object detection
- NLP for Document Analysis: TAPAS for query based tabular extraction to verify compliance with AICTE norms, and pdfplumber for document verification
- Pattern Recognition: Analyzed placement trends using Random Forest models and predict institutional compliance using XGBoost based on inspection scores
- Frontend & Backend Development: Built an interface using HTML, CSS, JS and backend powered by Django, PostgreSQL for structured data, and MongoDB for unstructured data storage