Problem Statement
The Project aims to establish a safe, secure, and empowering environment for citizens in public places by reducing crimes and enabling them from all backgrounds to work, travel, and reside in City without any risk of violence against them.
A core requirement is to apply data regression, correlation, and data mining methodologies to data that is available to provide specific crime patterns for each geographical region.
Value Proposition
Key elements of the solution include – data access, data management, analytics, and business intelligence through modular components & created single, integrated platform.
Various Statistical Models are built to –
- Hotspot & Trend Analysis
- Crime Analysis & Journey to Crime Identification
- Suspect Analysis & Repetitive Offenders Analysis
- GIS based Patrol Charts
Problem Statement
- Key Requirement for Image analytics Includes –
- Pre-emptive Analytical Surveillance
- Patrol Optimization
- Variance based activity detection
- Machine Learning-based Image Analytics
Value Proposition
- Image Recognition requires training of algorithms like CNN/RNN to detect image objects and faces
- Crowd-pooling & tagging along with camera trap capture tagging will be used to train models
- CNN/RNN models to be deployed to run image recognition for species tagging/ID tagging
- Image to be stored in BDL after tagging process is undertaken with repository update
Solution Strategy
Problem Statement
The System should be able to easily configure a chatbot with the requisite query/ response data feed on various schemes/ programs to enable interaction to enable seamless citizen engagements.
FAQs shall be easily embedded into the system. The system shall enable pre-defined smart answers to their general queries.
Value Proposition
A Chatbot is a Natural Language Processing based solution running on the query string execution.
Key Feature Includes –
- Pre-Suggested Keywords with a welcome screen
- Conversation Maturity using NLP Capability
- Autonomous Reasoning
- Predictive in nature
- Emoji based human like approach
- User can jump between Categories at any layer of hierarchy.
