DRISTEE — Navigation Assistant for Visually Impaired
DRISTEE: a real-time cloud-based navigation assistant for visually impaired users combining depth-aware mapping, object/face detection, and voice-command NLP.
What is ahead?
Scroll to watch it unfold. Each beat is one documented step of the project.
- 01
Capture
The camera frame is sent for cloud processing.
- 02
Object detection
Objects in view are detected.
- 03
Depth-aware mapping
Depth gives the distance and direction of each object.
- 04
Voice guidance
The result is spoken back to the user.
- 05
Outcome
The user hears what is ahead and how far away it is.
- 06
Familiar face
The user is told who is nearby.
- 07
Voice command
The spoken request is answered by the right module.
An illustration of how the system works, drawn from the write-up below. It is not recorded output.
The idea
DRISTEE (Depth-aware Recognition and Intelligent Scene Translation for Enhanced Engagement) is a real-time, cloud-assisted navigation assistant for visually impaired users, built under BRAC University's Research Seed Grant Initiative 2025. It fuses depth-aware scene mapping, object detection and facial recognition with voice-command NLP in a modular architecture, aiming to give a user personalized, low-latency spatial awareness — what's ahead, how far, and who's nearby — through natural spoken interaction rather than a screen.
Status
The project is ongoing, spanning model integration, implementation, and academic writing, and the research has now been submitted for journal publication. Method and evaluation details are held back until the review process concludes.
Tech stack & key skills
Core tools, methods and skills demonstrated in this project: