k
khatri_fahad

Fahad

@khatri_fahad

Computer Vision Engineer, YOLO and Real Time Deployment

Paquistão
Inglês, Urdu
Algumas informações são exibidas no idioma inglês.
Sobre mim
Computer Vision Engineer specializing in YOLO object detection, real-time deployment, and dataset quality auditing. I build pipelines in PyTorch, OpenCV, YOLO, and TensorRT, including multithreaded real-time systems running at deployment speed on consumer GPUs. Most detection models look strong on paper and fail on real data because the train/test split leaked. I audit for that first, so the numbers I report are the numbers you get. Recent work: PCB solder defect detection, real-time drone tracking on a TensorRT pipeline, and a 22,000 experiment LLM bias audit.... Saiba mais

Habilidades

k
khatri_fahad
Fahad
offline • 
Tempo médio de resposta: 7 horas

Conheça meus serviços

Visão Computacional
I will train and deploy custom yolo object detection models in real time
A partir deUS$ 90 /projetoMais detalhes
Anotação e marcação de dados
I will annotate images and video for object detection and segmentation datasets
A partir deUS$ 15 /projetoMais detalhes

Portfólio

Experiência profissional

Self_Employed

Computer Vision Engineer

Self Employed

Oct 2024 - Present • 2 yrs

Independent computer vision engineering across detection, deployment, and dataset quality. Built a two-stage PCB solder defect detection pipeline pairing a TensorRT-optimized YOLO detector with a downstream classifier, using custom IoU-based losses for small, low-contrast defects. Built a real-time FPV drone tracking system on a three-thread TensorRT pipeline sustaining deployment-speed frame rates on one consumer GPU. Ran a 22,000-experiment audit of hiring bias across five commercial LLMs. All work uses leakage-controlled dataset splitting and full metric reporting.