← Back to cohort

Saad Ahmed Bazaz

FAST · 2022 · I18 - 0621

—

Phone

—

GitHub

—

Academic

Program

BSCS

CGPA

—

Year

2022

Education

SEECS

Address

—

DOB

—

Career

Current role

—

Target role

—

Skills

Python, PyTorch, FastAPI, Django, ReactJS, Deep Learning, Machine Learning

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.

DeepDub 

DeepDub is an engine which can translate videos from one language to another - it dubs the human 
speech and also alters the lips of the actor so that it seems like they are speaking in the dubbed 
language. It comprises of a pipeline of Deep Learning and Machine Learning models which work 
together to produce the results. It is built to be modular from the ground up; it is easy to swap out 
modules for other modules, and to add more languages. To demonstrate the power of this engine, 
users can record themselves speaking some words and see themselves being dubbed, using the “Echo” app. To show the applications of the DeepDub in the industry, users can edit videos 
according to their desire, using the “Showup” app. 
Features include: 
- Extensive Language Support (German, Chinese, Urdu, Hindi, and Turkish, to English). 
- Useful side-products of this engine are that it can caption a video and also produce a translated 
subtitles file (.SRT file) 
- The Echo App (Live demonstration) 
- The Showup App (Video editing app with DeepDub integration) 









Technology Used: 
Python, PyTorch, FastAPI, Django, ReactJS 
Supervisor Name: 
Dr Mirza Omer Beg 
Group Members:   
AbdUrRehman Subhani (I18 - 0732) 
Saad Ahmed Bazaz (I18 - 0621)

AI enrichment

Saad Ahmed Bazaz is a BSCS graduate who contributed to DeepDub, a modular deep learning engine for video translation and lip-syncing. The project involved building a pipeline of ML models and developing frontend applications using ReactJS and FastAPI.

Skills (AI)

["Python", "PyTorch", "FastAPI", "Django", "ReactJS", "Deep Learning", "Machine Learning", "Video Processing", "Natural Language Processing"]

Status: ai_done

Provenance

Source file: Graduate Directory FAST School of Computing 2022 Final Version (07-06-2022).pdf
From job #25 page 224
Created: 1778170933