Expected grade: B
Actual grade: A- (Wut?)
Assessment and workload:
– 45% Oral examination. You read that right, oral examination!
– 55% Final examination, open book.
– This module is pretty interesting. It is an eye opening module which made me view visual image processing in a totally different viewpoint. Prof Ji introduces a lot of ideas in this module, and presents modern research results to us, most of them still fresh in recent years. The module begins with an introduction to frame theory which comes from analysis and has a lot of definitions. After the groundwork was laid, he moved on to talk about image processing, with the crown jewels of the module being the Unitary extension principle and the Restricted isometry property. Many ideas from cross disciplines and fields of study were also introduced, from optimisation algorithms, the same ones taught in MA5243, to computer science and machine learning concepts. Thus, people who take this can expect a lot of breadth but limited depth, although Prof Ji tries his best to go into greater detail on the proofs.
– Prof Ji is extremely knowledgeable in this field as it is his research area. I can tell that he teaches this module with quite a bit of passion as he shares with us modern research in this area. His digressions are very entertaining and it gets even better since I took this module with 2 other Singaporean friends and we have interesting discussions in class or after class about the use of image processing technology. However, the grouse is that he tends to erase what he writes too quickly before students can copy the interesting stuff he writes on the board.
– There are no tutorials for postgraduate modules. Prof Ji does not provide any questions either. He only gives us recommendations on reading some academic papers to read about the elaboration of the things he discussed in class.
– The oral exam was quite a scary affair. Prof Ji will summon you to his office and ask you 3 questions, he would give you time to think after he asks the question. My oral exam took me a grand total of 6-7 minutes, as I could not answer 2 out of 3 questions. The questions are non-standard questions that cannot be found in course material, and he claims that they’re common interview questions for people applying for R&D jobs in this field.
– Final exam was quite a strange paper consisting of 5 questions. I just modified the arguments presented in the lecture notes and other course material to answer the questions to the best of my ability. My friends and I got different answers for almost all the questions so we don’t know what’s going to happen in the final grade.
Head over to our Shop for more module content!