NIT-K Internship Project 2026

Developed under the initiative: "Creating Awareness in the Field of Artificial Intelligence through Hands-on Activities for Secondary School Children in Selected Districts of Karnataka, Kerala and Tamil Nadu."

Meet Whisper-bot

An offline, autonomous robotic system combining local Speech-to-Text (STT), Semantic Intent Classification, Local RAG Q&A using a Small Language Model (SLM), Computer Vision, an interactive LeNet-5 CNN visualizer, and smart Bluetooth Audio connectivity.

Core Capabilities

Whisper-bot combines computer vision and natural language processing for intelligent interaction

Face Detection

Advanced facial recognition with real-time detection and identification of multiple faces simultaneously

Instant Face Learning

Rapidly learns and remembers new faces with minimal training samples for personalized recognition

Natural Language Understanding and Instant Learning

Intelligent voice command processing with contextual understanding and semantic analysis. Learns new sentences instantly and applies it in real-time.

Object Detection & Tracking

Real-time identification and precise tracking of objects with continuous spatial awareness

LeNet-5 CNN Digit Visualizer

Interactive character recognition exposing intermediate layer topology (Input [28x28], Conv1 activations [8x24x24], Pool1 [8x12x12], Conv2 [16x8x8], Pool2 [16x4x4]) and 16x16 Fully Connected weight mapping (256 dense inputs to 10 logits).