Embed Size (px)
Citation preview
FRGS Review Workshop
Mohammad Faizal Ahmad Fauzi23rd January 2020
Multimedia University
• B. Eng. (Imperial College London, UK) 1997 - 1999 • PhD (University of Southampton, UK) 2000 - 2004
• Lecturer (MMU) 2004-2010• Senior Lecturer (MMU) 2010 – 2012• Associate Professor (MMU) 2012 – Present• Visiting Scholar (Ohio State) 2013-2014, 2017
• Grants: FRGS, ERGS, ScienceFund, TMR&D, IsDB, Fulbright, MCMC
• FRGS Evaluation Panel 2015 – Present
Speaker Background
2
Outline
FRGS Guideline 2020
FRGS Review Process
MyGrants Application
Example of Reviewers’ Comments
Q&A
FRGS Guideline 2020
Application
• Pemohon amatlah digalakkan untuk berkolaborasi dengan pihakIndustri/Agensi yang berkaitan bagi melaksanakan projek penyelidikantersebut (bukti dokumen adalah sekurang-kurangnya surat niat (letter of intent) dan lain-lain dokumen yang berkaitan).
• Pemohon amatlah digalakkan untuk mengemukakan bukti carian paten (patent search) bagi setiap permohonan projek.
• Setiap pemohon dibenarkan untuk mengemukakan hanya satu permohonanprojek sahaja pada setiap fasa.
• Permohonan baharu bagi mereka yang telah diluluskan FRGS pada fasa yang terdahulu akan diberi pertimbangan dengan bukti projek-projek tersebut telah75% selesai di dalam Modul Pemantauan Sistem MyGRANTS dan disahkanoleh RMC Institusi. Penyelidik dipohon untuk mengemaskini prestasi projekpenyelidikan secara atas talian melalui Sistem MyGRANTS pada setiap Februaridan Ogos setiap tahun.
• Tempoh perkhidmatan Ketua Penyelidik mestilah berbaki sekurangkurangnyadua (2) tahun di Institusi masing-masing mulai daripada tarikh tutuppermohonan FRGS.
National Priority Areas
Setiap permohonan hendaklah memenuhi salah satu Sektor Keutamaan Negara (National Priority Areas - NPAs) seperti berikut:-
a) Jaminan Makanan (Food Security);b) Jaminan Tenaga (Energy Security);c) Tanaman Perladangan (Plantation Crops);d) Keselamatan Siber* (Cyber Security);e) Jaminan Air (Water Security);f) Kepelbagaian Bio (Biodiversity);g) Penjagaan Kesihatan dan Perubatan (Healthcare and Medicine);h) Alam Sekitar dan Perubahan Iklim (Environment and Climate Change); ataui) Pengangkutan dan Mobiliti (Transportation and Mobility)
*Nota: Sektor Keutamaan ini adalah merangkumi IR4.0 (Cyber Security, Big Data, Artificial Intelligence) yang merujuk kepada Dasar Negara Mengenai Industri 4.0 (National Policy on Industry 4.0).
Research Clusters
Setiap permohonan hendaklah memenuhi salah satu klusterpenyelidikan seperti berikut:-
a) Sekuriti dan Keselamatan Makanan (Food Safety & Security);b) Infrastruktur Asas (Basic Infrastructure);c) Perubahan Iklim dan Alam Sekitar (Climate Change & Environment);d) Kesihatan (Health);e) Pendidikan dan Masyarakat Madani Celik Ilmu (Education &
Knowledgeable Civil Society);f) Keselamatan Negara (National Security);g) Kesejahteraan Sosial dan Ekonomi (Social & Economic Wellbeing);
atauh) Teknologi Terkehadapan & Pembuatan Termaju (Frontier
Technologies & Advanced Manufacturing)
Research Domain
Setiap permohonan hendaklah memenuhi salah satu domainpenyelidikan seperti berikut:-
a) Sains Tulen dan Gunaan (Pure and Applied Sciences);b) Teknologi dan Kejuruteraan (Technology and Engineering);c) Sains Kesihatan dan Klinikal (Clinical and Health Sciences);d) Sains Sosial (Social Sciences);e) Sastera dan Sastera Ikhtisas (Arts and Applied Arts);f) Warisan Alam dan Budaya (Natural and Cultural Heritage);g) Teknologi Maklumat dan Komunikasi (Information and
Communication Technology).
Output
• Talent
– 1 PhD or 2 Masters for 3-year project
– 1 Master for 2-year project
• Publications
– At least 2 indexed journals
– Must acknowledge FRGS in the paper
• Intellectual Properties
– Encouraged to apply for IP protection
Evaluation Criteria
• Penyelidikan yang dipohon mestilah merupakanpenyelidikan yang dapat menghasilkan suatuidea/teori/konsep/kaedah/model/proses baru; dan
• Penyelidikan yang dapat menambahbaik sesuatu polisi, metodologi dan model penyelesaian yang sedia ada; atau
• Penyelidikan yang merangkumi isu-isu kemanusiaandan kemasyarakatan bagi tujuan peningkatan nilaikehidupan dalam negara dan sejagat; atau
• Penyelidikan yang berpotensi untuk menyumbangkepada agenda strategik negara.
FRGS Review Process
Role of Evaluators
• Internal Evaluation:
– Department / Faculty / Institute level
– University level
Internal evaluators improve the quality of grant submissions
• IHL Panel requirements:
– 2 internal reviewers (selected by IHL)
– Decision: Highly recommended OR Rejected
Note: only highly recommended decision will be considered for national level FRGS evaluation.
• KEY CHALLENGE : How to improve?
Role of Evaluators
• External Evaluation (national level):
– According to clusters
• Engineering always has the largest submissions
– Sponsor level
– Other constraints
• Total numbers
• Grant allocation
• Etc.
• KEY CHALLENGE : How to reject?
MMU Success Rate
• 2019– Submitted 355– Approved 53 (14.9%)
• 2018– Submitted 146– Approved 21 (14.4%)
• 2017– Submitted 124– Approved 14 (11.3%)
• 2016– Submitted 103– Approved 30 (29.1%)
• 2015– Submitted 129– Approved 35 (27.1%)
2019 submission nationwide: 7122
2018 submission nationwide: 7079
2017 submission nationwide: 6284
2016 submission nationwide: 4486
2015 submission nationwide: 3622
MyGrants Application
Title
Title
Title
• E.G.– Novel Methods for Spotting and Recognition of Spontaneous
Facial Micro-Expressions from Naturalistic Situations
– Grading and Histoscoring of Breast Carcinoma: Bringing Machine Intelligence to Malaysian Histopathology Research
– Practical Interference Alignment Techniques with Reduced CSI Dependency for 5G and Beyond Wireless Communications
– Landslide Prediction with Wireless Sensor Networks using Non-linear Learning Models
– Intelligent Design of Real-Time Scheduling in Stochastic Environment of Flexible Flow Shop
– Modified Principal Component Analysis on Hamming Cube for Action Recognition
Research Information
Track Record
Track Record
Track Record
Executive Summary
Executive Summary
Executive Summary/Abstract
A micro-expression is a brief and involuntary facial movement which reveals a genuineemotion that a person tries to conceal. While psychologists have been studying facial micro-expressions since the 1960’s, only until the last few years, the possibility of spotting andidentifying these micro-expressions automatically using machine algorithms has arisen. Asopposed to normal facial expressions which have been intensively studied for decades, this isa contemporary field of research; we are among three leading institutions in this new field ofresearch. The analysis of spontaneous micro-expressions has many potential applications;clinical diagnosis of psychological conditions such as autism and depression, and also criminalinterrogation and lie detection.
This project aims discover novel methods for automatic spotting and recognition ofspontaneous facial micro-expressions. Firstly, this project endeavors to create the first-everdataset for spontaneous micro-expressions elicited from naturalistic situations to challengethis growing research field. Existing datasets were constructed using only emotional videoclips as visual stimuli. Secondly, the crucial first task of micro-expression spotting demands fornovel algorithms that can effectively locate the onset, offset and apex frames from acontinuous video to a good level of precision. Current heuristic-based methods for measuringtheir occurrences do not generalize well across different persons. Finally, present state-of-the-art recognition performance can be improved with new spatio-temporal featurerepresentations that can robustly amplify and discriminate between micro-expression signalsfrom different categories.
The expected output of this project is three-fold in terms of new fundamental knowledge andfindings: Algorithms for representing and spotting micro-expression occurrences, theoreticalspatio-temporal models for better recognition of micro-expressions, and valuable know-howinto the construction of spontaneous micro-expression dataset from naturalistic scenarios.This project will have far-reaching impacts towards the well-being of the society and publicsafety of the nation with its strong relevancy to governmental policies.
Problem statement
Objectives
Research methodology
Expected outcomes
Significance of output
Example
Research Background
Research Background
Problem Statement
A micro-expression is a brief and involuntary facial movement which reveals a genuine emotion that a persontries to conceal [1, 2]. The duration of a micro-expression is in the range of 1/3 to 1/25 seconds (the generallyacceptable upper limit is 1/2 second at most [2] ), and it typically appears with low intensity, almost difficult tobe observed with the naked eye. As such, a person exhibiting micro-expressions is usually unaware or unable tocontrol it spontaneously through willpower. While psychologists have been studying facial micro-expressionssince the 1960’s, only until the last few years, it has begun to garner more attention in affective computingstudies. Machine recognition of general facial expressions has been intensively studied in the field of computervision for decades [3], but very little research has been done to process and analyze micro-expressions usingcomputers. The analysis of spontaneous micro-expressions has many potential applications; particularly forclinical diagnosis of the emotional well-being of the patient, and various mental/psychological conditions such asautism and depression; and for criminal interrogation, where micro-expressions formed essential clues fordetecting lies and cheating among suspects.
Analysis of spontaneous facial micro-expressions is very challenging, even for human beings [1], due to theirshort duration and subtleness in intensity. Generally, given a continuous video clip of a person-of-interest,analysis of micro-expressions includes performing two inter-connected tasks: “spotting”, followed by“recognition”. A recent work collecting spontaneous micro-expressions [4] pointed out that human annotatorsfound the task of spotting challenging and heeded the advice of Ekman to first view the video frame-by-frameand then with increasing speed [5]. Ekman’s study [6] shows that people without training were only able toperform ME recognition slightly better than chance, while a trained psychologist is able to perform better butonly up to an accuracy of 40+%. This was not a considerable obstacle for detecting and recognizing normal or“macro” expressions. Hence, there are several challenges in this relatively new field of work that is worthpursuing, particularly on the state of current data, and the dual tasks of ME spotting and recognition.
Example
Problem Statement
Elicitation of MEs is difficult because it is only present at specific situations, when a person tries to suppress feltemotions but fails to do so, resulting in a leaked emotion manifested as micro-expressions [2]. Both current state-of-the-art datasets, i.e. SMIC [4] and CASME II [7] (CASME [8] is the predecessor of CASME II), were collected by asking subjectsto watch emotional video stimuli while attempting to keep neutralized faces within a “high-stakes” situation (they will bepenalized for any emotion leaks). However, facial MEs in more naturalistic situations should not be limited to thisparticular paradigm of elicitation. Realistic scenarios such as face-to-face conversations/interviews, or a high-stakesituation involving telling of lies about what they have read or seen previously, may also exhibit MEs. Moreover, thesesituations reside in a more unconstrained environment (than in a very constrained studio setting) that requires dealingwith other irrelevant facial movements. Also, both the SMIC and CASME II datasets were constructed by extracting shortvideo samples containing a single emotion each. This is inadequate to cater for a complete spot-then-recognizeprocedure, as naturalistic situations may typically comprise of a number of micro-expressions, among a spate of otherkinds of movements.
In order to recognize what type of MEs are present, the MEs must first be spotted, or detected in the video. Thus, thecapability of recognition is strongly dependent on the success of correctly spotting the MEs. Manual spotting of the onset(starting), apex (strongest) and offset (ending) frames in video is often a time-consuming task that requires highreliability. Compared to methods for recognition, spotting is also a significantly understudied problem in this area.Existing approaches are fraught with many impediments, i.e. optimal threshold parameters are overly sensitive and hardto identify [9], the presence of eye-blinks which had to be manually removed [10]. This is clear indication that research inmicro-expression spotting is still at an early stage.
While the original recognition baselines of CASME II and SMIC starting off (in 2014) at an F1-score of 0.35 and 0.40respectively, the latest state-of-the-art methods [11] and [12] (our group’s latest work) are now achieving a top score of0.57 and 0.60 for CASME II and SMIC respectively. Nevertheless, this is still far from being reliable enough for use inautomated recognition systems. Interestingly, many methods in literature have worked primarily on two separate aspectsof the recognition pipeline: temporal pre-processing [12, 13] and spatio-temporal feature representation [14, 15] withoutexploiting how they can be seamlessly integrated. On ME data, traditional feature representations [16] do not fare aswell as with visual data from other domains such as normal facial expression and human actions/activities. It is likely thatsubtle facial changes across time are difficult to capture and encode into meaningfully discriminative information.Moreover, the use of motion magnification has not been thoroughly explored; preliminary attempts were made [17, 18]but improvements had been limited.
Example
Hypothesis
Due to the above-mentioned problems and challenges, we hypothesize that facial micro-expressions will be a fast-growing field of study in computer vision and affective computing in the coming years. Research on normal facialexpressions have reached a saturated stage with off-the-shelf products such as Affectiva [19] and Emotient [20]able to attain almost perfect emotion/expression recognition rates. Meanwhile, research on subtle micro-expressions remains a new challenge. As of now, there are only three research groups in the world who have beenactively working in this new area since 2014 – namely Oulu University, the Chinese Academy of Sciences (CAS) andMultimedia University (MMU) (our group).
The recent emergence in advanced techniques such as efficient dense optical flow [21] and Eulerian motionmagnification [22] presents viable solutions for subtly changing facial motions. New spatio-temporal encodingmethods [23] that are robust against noise can be potentially explored and innovated further for this domain ofwork. It is envisaged that machine-assisted analysis of facial micro-expressions will in the foreseeable future,provide auxiliary support for social and scientific domains such as forensic science, criminal investigation, medicaldiagnosis and psychotherapy.
Example
Research Questions
The direction of this research can be summarized with the following questions:
1. How can a new micro-expression dataset be acquired from a comprehensive range of naturalistic scenarios to accomplish automatic spot-and-recognize procedure?
2. How can facial micro-expressions (in terms of onset, offset and apex frames) be spotted more precisely to better construct micro-expression sequences?
3. How can robust spatio-temporal representations be designed to recognize the spotted facial micro-expressions more effectively?
4. How do machine algorithms fare against humans in the task of spotting and recognizing micro-expressions?
Example
Relevance to Government Policy
In the past decade, the European community has been paying close attention to the transformation of societythrough their very own Social Signal Processing Network (SSPNET) [31] for "social intelligence" research, a multi-disciplinary foray into how social signals such as nonverbal behavioral cues such as emotions, sentiments, gesturesand prosody can help understand society and interaction between humans in a better way. We take cue of suchimportance from developed nations, to enable research in our nation to transform society as well.
With the fundamental nature of the proposed project, the expected research outcomes have deeply-rootedconcepts that are directly transferable to various applied domains, and have far-reaching impacts on the well-being of the society and public safety of the nation. The prediction of human emotional states, especially insituations where patients attempt to withhold their emotions is particularly beneficial for clinical assessment ofpsychological conditions such as autism, schizophrenia and depression. This is in line with Malaysian government'spolicy under the Healthcare National Key Economic Areas (NKEA), which aims to create a supportive ecosystem forclinical research and diagnosis.
In the aspect of criminal investigation and forensic science, our contributions enable investigators to see throughthe hidden emotions of deceptive suspects and criminals, who may attempt to conceal their actual agenda andmotives. Hence, this project has strong relevancy in terms of its potential contribution towards supporting theNational Key Results Areas (NKRA) under Governmental Transformation Policy (GTP) 2.0 in two priority areas: (I)Fighting Corruption and (II) Reducing Crime. Interrogations of suspects conducted by police or other authorities,and drug addict rehabilitation efforts can be greatly enhanced with the help of such technology.
Example
Objectives
Objectives
Objectives
This research study embarks on the following objectives:
1. To create the first known dataset of spontaneous facial micro-expressionselicited from a wide range of challenging naturalistic situations, complete withcomprehensive emotion labels and coding.
2. To formulate new micro-expression spotting algorithms that can effectivelylocate the onset, offset and apex frames from a continuous video to a goodmeasure of precision.
3. To propose new spatio-temporal feature representations that can robustlyencode micro-expression signals to further improve recognition performance.
4. To validate the machine-automated unified spot-and-recognize scheme againsthuman performance.
Example
Methodology
Methodology
Methodology
# of Research Activities should be more than # of Milestones!!
Methodology
Example of Flowchart
Expected Output
Expected Output
Expected Output
Expected Output
Impact on Society, Economy and Nation
The well-being of the society and public safety of the nation is a key milestone in ourcountry’s pursuit towards becoming a first-world, developed nation. This project willhave a far-reaching impact as a fundamental research on the prediction of humanemotional states, which is beneficial to various applied domains, especially in the areaof healthcare and crime prevention. This project has a strong relevancy in terms of itspotential contribution towards supporting the Healthcare NKEA, and GovernmentalTransformation Policy (GTP) 2.0 in two priority NKRAs: (I) Fighting Corruption and (II)Reducing Crime.
Novel theories/New findings/Knowledge
1. Fundamental knowledge into how micro-expression movements can be characterized and spotted from a long continuous video
2. Novel theoretical models that can robustly encode subtly-changing micro-expression sequences for better recognition performance.
3. New knowledge and know-how into the construction of a spontaneous facial micro-expression dataset (the very first in Malaysia) from naturalistic situations.
Example
Quality of Proposal
Equipment & Material
Budget
Budget
Example of Reviewers’ Comments
1. A Good Fit
2. Communication – Clear & Concise
3. A Solid Idea or Conceptual Analysis
4. Work Plan or Methodology
FRGS Reviewer’s Message
FRGS Reviewer’s Message
1. A Good Fit
– Review the mission of the Ministry of Education and the guidelines of the specific grant program to which the applicant is applying. To make project a good fit.
• E.g. of comments
– Insufficient relevance to the development of ..[area].. applications
– Equipment is ill-suited for a multi-user environment
– Insufficient justification in terms of number of researchers/GRAs
FRGS Reviewer’s Message
2. Communication – Clear & Concise Language is key to a good proposal
– Reviewers have limited time and energy to devote to the proposal. If the applicant makes them work too hard just to understand what he/she is trying to say, they will have little energy left to assess the applicant’s ideas.
– Good ideas can be overlooked if the reviewer becomes distracted by typos, poor grammar, or other errors in construction.
FRGS Reviewer’s Message
• E.g. of comments (cont.)
– Proposal is poorly written (confusing, not logical, poorly organized, typos, etc.)
– Lack of detail
– Lack of a thorough thoughts process and clear story; lack of focus
– Improper or inadequate literature citations
– Improper use of certain terms
– Making assumptions/statements without having a citations or preliminary data to back it up
FRGS Reviewer’s Message
3. A Solid Idea or Conceptual Analysis– A strong research proposal takes an innovative idea and
shapes it into a coherent plan that includes sufficient data and references.
– PLs often fail to provide detail on the thought process behind the experimental plan.
– The proposal should also include contingency plans to address potential obstacles.
• E.g. of comments– Objective needs to be more focused– Project is too ambitious (unrealistic goals)– Not sure what new insights will be gained from the proposed
research– Lack of supporting preliminary data to demonstrate feasibility
FRGS Reviewer’s Message
• E.g. of comments (cont.)
– Proposal does not address obstacles encountered by others
– Plan is illogical
• Successive goals rely on unproven assumption
• Sticks to a narrow hypothesis without discussing or making allowances for alternative possibilities or approaches
• Too big a leap from preliminary data to the proposed hypothesis; failure to provide sound scientific data for the support of the hypothesis
• Idea is a solution in search of a problem
FRGS Reviewer’s Message
4. Work Plan or Methodology– Reviewers look for a logical, well-designed work plan that
includes sufficient detail to describe:– How, when, and by whom the work will be done– How the data will be analyzed, and– That the necessary skills and resources are available to
complete the proposed research
• E.g. of comments– Serious flaws in the experimental design; additional
experiments are required– Using sub-optimal technique: i.e., using out-of-date
techniques, or conversely, using unnecessary new techniques when standard techniques will work just as well
FRGS Reviewer’s Message
• E.g. of comments (cont.)
– Equipment is ill-suited for the proposed projects
– Essential piece of equipment or system not requested/available
– Timeline is overly optimistic
– Failure to provide a reasonable plan for exchange of information and coordination among investigators
– Lack of expertise in a specific area of the research (usually with respect to technology/protocol)
– Long-term equipment maintenance plan is insufficient
Other Comments
• It is not clear what question is being addressed by the proposal.
– In particular, it is not clear what the outcome of the research might be, or what would constitute success or failure. It is vital to discuss what contribution to human knowledge would be made by the research.
• The question being addressed is woolly or ill-formed.
– The committee are looking for evidence of clear thinking both in the formulation of the problem and in the planned attack on it.
• It is not clear why the question is worth addressing.
– The proposal must be well motivated.
• There is no evidence that the proposers will succeed where others have failed.
– It is easy enough to write a proposal with an exciting-sounding wish-list of hoped-for achievements, but you must substantiate your goals with solid evidence of why you have a good chance of achieving them.
• The proposed research has already been done - or appears to have been done.
– Rival solutions must be discussed and their inadequacies revealed.
59
Other Comments
• A new idea is claimed but insufficient technical details of the idea are given for the committee to be able to judge whether it looks promising.
– Since the committee cannot be expert in all areas there is a danger of overwhelming them with technical details, but it is better to err by overwhelming them than by underwhelming them. They will usually get an expert referee to evaluate your idea.
• The proposers seem unaware of related research.
– Related work must be mentioned, if only to be dismissed. Otherwise, the committee will think that the proposers are ignorant and, therefore, not the best group to fund. The case for support should have a list of references like any paper, and you should look at it to check it has a balanced feel - your referee will do so. Do not make the mistake of giving references only to your own work!
60
Other Comments
• The proposal is badly presented, or incomprehensible to all but an expert in the field. – Remember that your proposal will be read by non-experts as
well as (hopefully) experts. A good proposal is simultaneously comprehensible to non-experts, while also convincing experts that you know your subject. Keep highly-technical material in well-signposted section(s); avoid it in the introduction.
• The proposers seem to be attempting too much for the funding requested and time-scale envisaged. – Such lack of realism may reflect a poor understanding of the
problem or poor research methodology.• The proposal is too expensive for the probable gain.
– If it is easy to see how to cut the request for people/equipment/travel, etc. to something more reasonable then it might be awarded in reduced form. More likely, it will be rejected.
61
Common Failings
1. Did not follow application guidelines exactly.2. Not new or lack of original ideas
- e.g. routine application of known techniques.
3. Lack of knowledge of published relevant work4. Diffuse, superficial or unfocused research plan5. Lack of experience in the essential methodology
- uncritical approach
6. Insufficient experimental detail- or questionable reasoning in experimental approach
7. Unrealistically large amount of work8. Unacceptable scientific rationale9. Uncertainty concerning the future directions
62
Messages From RMC
Source: MOHE 2018
Budget
Vote 11000 Labour GRA (Pursuing Master: RM1800,
Pursuing PhD: RM2300)
Vote 21000 T & T Conference & training: 20%
Total T & T: 40% of project cost
Vote 24000 Rental Relates to research
Vote 27000 Research
Materials
Relates to research
Vote 28000 M & M Minor repairs & modifications
Vote 29000 Special
Services
5% of project cost
Vote 35000 Accessories
& Equipment
40% of project cost
TOTAL RM250,000.00 (Maximum)
3-Year (Maximum)
FRGS Guidelines
• Pemohon amatlah digalakkan untuk berkolaborasi dengan pihakIndustri/Agensi yang berkaitan bagi melaksanakan projek penyelidikantersebut (bukti dokumen adalah sekurang-kurangnya surat niat (letter of intent) dan lain-lain dokumen yang berkaitan).
• Pemohon amatlah digalakkan untuk mengemukakan bukti carian paten (patent search) bagi setiap permohonan projek.
Questions?
