← Back to cohort

Sher Ali

FAST · 2019 · i19-2007

—

Phone

—

GitHub

—

Academic

Program

BSCS

CGPA

—

Year

2019

Education

SEECS

Address

—

DOB

—

Career

Current role

—

Target role

—

Skills

Pytorch, Python, OpenCV, Flask, Pycharm, Github

Verbatim text

The exact text the LLM saw on the page (or the booklet text from the old import). This is what powers semantic search.

Deep Portrait Drawing
The aim behind developing Deep Portrait Drawing is that Every day, lots of crimes take place everywhere around the world but the police rarely catch the culprits. Police try to get the description of the culprit from eyewitnesses, but the problem with this is that not every police station has an expert sketch artist. This requires the artist to have years of experience in drawing people’s faces. Even if a professional artist is there, it eventually results in a very hard constrained sketch. Furthermore, this process is time-consuming and expensive. DeepPD will help users to draw rough or sparse sketches and get realistic looking images in return that would be helpful in solving the above problem.
Features include:
- A user will be able to draw and erase the sketch on the canvas using different size brushes
- User will be able to convert his/her sketch into a realistic portrait image.
- User can download/export the generated image on his/her PC.
- User can continue editing the sketch or finish the sketch after which the user will not be able to change the sketch
- Users can see instructions about how to use the software i.e. how to sketch the portrait using different brushes and convert the sketch to the realistic looking portrait image
DEEP PORTRAIT DRAWING
OVERVIEW
A Novel Deep Learning model which takes free hand sketches and generate realistic human face images
ARCHITECTURE
Component Embedding Module
Feature Mapping Module
Image Synthesis Module
TIMELINE
SEP - OCT Filtered dataset along with edge maps
FEB - MAR Combined feature maps from feature vectors
NOV - DEC Segmented features by Component Embedding
APR - MAY Synthesized image & cross platform application
Group Members
Hassan Saif 19i-0560
Faizan-Ul-Hassan 19i-1976
Sher Ali 19i-2007
Supervisor Dr. Akhtar Jamil
Co-Supervisor Dr. Uzair Iqbal
Technology Used:
Pytorch,Python,OpenCV,Flask,Pycharm,Github
Supervisor Name:
Dr Akhtar Jamil
Group Members:
Hassan Saif (i19 - 0560)
Faizan ul Hassan (i19 - 1976)
Sher Ali (i19 - 2007)

AI enrichment

Sher Ali is a BSCS student who contributed to a group project developing a deep learning application for generating realistic portraits from sketches. The project utilized PyTorch, Python, and OpenCV to create a cross-platform tool for law enforcement sketch enhancement.

Skills (AI)

["Python", "PyTorch", "OpenCV", "Flask", "Deep Learning", "Image Synthesis", "Git"]

Status: ai_done

Provenance

Source file: FAST - School of Computing -Graduate Directory-2023.pdf
From job #14 page 295
Created: 1778112745