Search Contact information
University of Cambridge Home Professor Roberto Cipolla, Department of Engineering


Guidelines on PhD research and supervision

Roberto Cipolla, August 1995

Formal guidelines

The Board of Graduate Studies set out the requirements of the PhD dissertation as follows:

"The dissertation must be clearly written, must take into account previously published work on the subject and must represent a significant contribution to learning (eg discovery of new knowledge, connection of previously unrelated facts, development of new theory or revision of old views.)"
It must include an abstract which clearly states the nature and purpose of the investigation and a bibliography which makes adequate reference to the relevant literature.

Skills that have to be developed

To achieve the PhD degree it is necessary to demonstrate that you have mastered the skills necessary to carry out research to professional standards. The point of the PhD is not to demonstrate your brilliance (although this might also occur), but to demonstrate that you have mastered a set of research skills.

Professional research standards means:

  1. You have something to say: i.e. you are able to present a coherent argument and can tell a story that your world-wide peer group is interested to hear (the thesis).
  2. You are able to evaluate the worth of what others are doing. Literature surveys should demonstrate that you have the maturity, critical and analytical skills to compare your work to previous and contemporary research and to point out the limitations.
  3. You have the astuteness to discover where to make a contribution and the ability to evaluate and re-evaluate your contribution.
  4. You can communicate effectively to the world wide peer group by writing clear, precise, logical conference and journal articles and making presentations at international conferences, workshops and seminars. You can demonstrate the importance/interest of your research to expert and non-expert visitors.
  5. You have mastered the appropriate experimental, mathematical and computational research skills. You are able to conduct literature searches, review conference and journal submissions.
  6. You are able to formulate plans to meet short-term and long-term goals. You are able to meet deadlines. Professional means that you have the determination and application to work to the conclusion of what you set out to do. Plans and goals will of course change but make sure you address the underlying reasons.

The dissertation

The PhD dissertation aims to allow the examiners to judge whether the candidate has met the above requirements. It should not be a record of all of the student's work in the department.

The dissertation should be structured to include:

  1. Identification of unsolved problem and reason for solving it. The nature and purpose/motivation for the investigation should be clearly stated. The thesis approach/standpoint and whether the purpose was substantially achieved should also be made clear.
  2. Status of research in direction of solution. The relevant background material and limitations of existing methods. The candidate must show that he/she has an adequate knowledge of the subject and of the literature and can critically place his/her work in a wider context. The literature survey should not be encyclopaedic.
  3. Find a solution. Development of own ideas and theoretical framework backed with mathematical analysis.
  4. Demonstration that it is a solution. This should involve the implementation, justification for assumptions and evaluation of evidence. Demonstrate analytical skills.
  5. Assess the suitability of the solution
  6. Evaluate the importance of the contribution
  7. Identify directions for future work
  8. Appendices
  9. Complete bibliography with numbered list of references. References in text should use names and numbers.

My approach to supervision

  1. Give a small initial project in the first year with a definite deadline before April of the first year. This will highlight problems in approach and if conducted to conclusion will lead to a conference paper (typically BMVC).
  2. Make sure students are familiar with the professional standards of published work by getting them to read one good PhD dissertation and getting students to regularly review literature and attend reading group.
  3. All students encouraged to present at a group seminar, national and international conference so that they face external criticism of research and hone presentational skills.
  4. Encourage independence and transfer dependence after the first year. Make sure students are able to set their own goals and meet deadlines. Make students understand underlying problems when they fail to meet deadlines.
  5. Ensure that students realise that original research contributions come about after diligent review, concentration and analysis. The world is your oyster and in research everything is up for grabs.

    'Caminante no hay camino. Se hace camino al andar'' (Traveller, there are no paths. Paths are made by walking)'

  6. Ensure students have a professional attitude to research. Present legible written work at mutually convenient deadlines. Attend research meetings with notes on research in progress methods attempted and papers read. Treat research like any other job and maintain average though flexible working hours (minimum 6hrs real work/day). Students who fail to work on average 9-5, five days a week almost always fail to make satisfactory progress and overrun.
  7. Establish a peer group to advise and give comments. Ensure that second and third year students help in the training of newcomers. Ensure that students make a modest contribution to group activities
  8. Students who fail to make adequate progress by time of first year report (May) are encouraged to resubmit by September. Failure at the second report stage will lead to termination of funding.
  9. Final year students to give a group seminar before writing dissertation to help organise story line and structure.
  10. Effective communication skills require the student to be able to write concisely, logically and in grammatical english. These skills are developed by reading journal articles and in discussion with colleagues. You are also encouraged to read as widely as possible outside your discipline and to discuss your research ideas with friends and non-experts. Illustrations can lead to dramatic improvements in the effectiveness of papers.

What students can expect from supervisor

  1. Early direction. Initial project formulation and plan.
  2. Exposure to all elements of research training.
  3. Advice on reformulation of goals and plans
  4. Positive feedback on research proposals. Interest in what they are doing and why.
  5. Criticism if fail to meet professional research standards.
  6. Meetings as requested by student and supervisor.
  7. Support and encouragement in time of research, financial or personal crisis.

What is a sucessful outcome?

  1. PhD dissertation of quality on time (3 years)
  2. An academic journal article (eg IJCV, PAMI or IVC)
  3. Oral presentation at an international conference (eg ICCV, ECCV, CVPR)
  4. Meeting, arguing with and impressing professional researchers at workshops, seminars and visits.
  5. Stimulating experience for student and supervisor.
  6. Benefit to PhD peers in ideas, software, research grants or demonstrations. Research project is passed on to the next generation of students.
  7. A job at the end of three years.

Postgraduate Information

Copyright © 2008 Roberto Cipolla