← Muhammad Hasheem Iftikhar
The Bridge of Hopes · 2024 – 2025

Assistive AI for Children with ADHD

A new Urdu alphabet dataset and classifier for learning support that adapts to each child

Muhammad Hasheem Iftikhar, AI Research Team Lead

The Bridge of Hopes AI team · research initiative led by Dr. Anam Qureshi

The Bridge of Hopes AI team Team lead selection Homepage
Urdu alphabet letters from the dataset's 39 classes
Urdu alphabet dataset. A new dataset of Urdu letter images across 39 classes, built because no public equivalent existed.
6,220images collected
39Urdu letter classes
95.52%CNN accuracy
0.96F1 score

Goal

Tutoring that fits how each child learns

The Bridge of Hopes builds AI assistive technology for children with ADHD and special needs. The aim is a system that recognizes how each child learns and tutors them accordingly, in their own language.

I set the research direction by studying where ADHD traits meet AI capabilities, so the system could identify how each child learns and tutor them accordingly.

Work

From no data to a working classifier

  1. Research direction

    Studied where ADHD traits meet AI capabilities to decide what the system should learn about each child.

  2. Urdu alphabet dataset

    Built a 6,220-image, 39-class dataset of Urdu letters, a resource with no public equivalent.

  3. Letter recognition

    Directed a CNN classifier that reached 95.52% accuracy (F1 = 0.96) on the dataset.

My role

Founding researcher to team lead

Mar 2024

Founding AI researcher

Selected through a competitive application and joined before the project had funding. The groundwork we laid helped the initiative secure funding later.

2025

AI Research Team Lead

Appointed AI team lead by Dr. Anam Qureshi on my seniors' recommendation.

More research

ACL submission · in preparationEduMem ACM FAccT 2027 · under reviewWhen AI Work Becomes Official Thesis · MICCAI 2027 plannedEnhancing mAIstro