Skip to content
#

pakistandataset

Here are 7 public repositories matching this topic...

Language: All
Filter by language

Benchmarking Vision-Language Models (VLM) for Visual Impairment Assistance in Pakistan using real-world local environments. This project evaluates VLMs like LLaVA-13B and introduces a human-centered, VI-oriented description format focused on navigation, spatial awareness, safety, and accessibility.

  • Updated May 19, 2026
  • Python

Add this topic to your repo

To associate your repository with the pakistandataset topic, visit your repo's landing page and select "manage topics."

Learn more