2 GRA Vacancy for Automated Detection of Visual and Speech Contents for Film Censorship using Deep Learning Project
|University / Institute: Multimedia University|
|Research Center/Lab (optional):|
Automated Detection of Visual and Speech Contents for Film Censorship using Deep Learning
|Start Date 2 Jan 2019|
|End Date: 30 Nov 2020|
|Funding Agency: Telekom Malaysia R&D Grant|
|Contact Email(s): firstname.lastname@example.org|
|Nationality Requirements on Candidate(s): Local or Foreigner|
|Research Scholar 1|
Monthly Salary: RM2500 (To pursue Masters full time) (depending upon candidate’s experience).
Main Task: Data collection, algorithm development and simulation for audio censorship.
Main Requirement: Bachelor’s degree with honours in Electronics/Electrical/Computer Engineering/Computer Science discipline, preferably 2nd class or above.Research Scholar 2
Monthly Salary: RM3000 (To pursue PhD full time) (depending upon candidate’s experience).
Main Task: Data collection, algorithm development and simulation for video censorship.
Main Requirement: Master’s degree with honours in Electronics/Electrical/Computer Engineering/Computer Science discipline, preferably 2nd class or above.
Location: Faculty of Engineering, MMU Cyberjaya
Benefits of the project (not limited to):
-Attending local conference(s) based on oneself effort in publication.
• To evaluate public nudity scene video dataset and prepare offensive /foul language data set.
• To implement and evaluate existing techniques on public data set and newly created offensive audio dataset. To do testing on video/audio data provided by UNIFI TV.
• To perform coding using Python.
• To develop, train and evaluate CNN for nudity scene detection in video stream.
• To develop, train and evaluate RNN to detect foul languages in video.
• To do any relevant administrative work/purchasing matters as requested.
• Research and publication work.
• Good knowledge of image and video processing.
• Good English proficiency.
• Self-motivated, requires minimal supervision, resourceful, keen to learn, possess good communication skills and able to work under pressure.
• The candidate ideally should be proficient in Python scientific programming. This project uses the deep learning technology for video and audio classification. Therefore, candidate that have completed the course Practical Deep Learning For Coders, Part 1 at
http://course.fast.ai/ will have an added advantage.
• Deep learning knowledge will mainly be evaluated based on previous completed projects
You may refer to the link below for sample advertisement on gra vacancies