Recognition of facial expressions - Knowledge Based Systems ...

More documents

Recommendations

Info

--------------------------------------------------------------- 1: P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 Exp AU1 AU2 AU4 AU5 AU6 AU7 AU9 AU10 AU11 AU12 AU15 AU16 AU17 AU18 AU20 AU21 AU22 AU23 AU24 AU25 AU26 AU27 2: 3 4 3 5 1 1 3 2 3 5 Fear 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 0 0 3: 4 5 5 7 4 3 3 5 6 2 Surprise 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 4: 2 4 2 5 3 1 3 2 2 3 Sadness 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 5: 2 3 2 3 1 1 3 2 2 2 Anger 0 0 1 0 0 1 0 0 0 0 0 0 1 0 0 0 0 1 1 0 0 0 ............................................................. 483: 3 3 3 6 4 2 2 6 6 2 Surprise 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 484: 3 3 2 2 1 1 2 1 2 6 Happy 0 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 485: 3 3 2 4 1 1 2 1 2 6 Happy 0 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 //--------------------------------------------------------------- Listing 8. Final data extracted from the initial database (version II) For every row, the sequence of Action Units was encoded by using the value “1” for denoting the presence of the AU and “0” for absence. The discretization process for the given example was done on 7 classes per parameter basis. For the conducted experiments, there were also generated training files with 5 and 8 classes per parameter discretization. CPT Computation Application The application is used for determining the values in the table of conditioned probabilities for each of the parameters included in the Bayesian Belief Network. Initially, a BBN model can be defined by using a graphical-oriented application, as GeNIe. The result is a “XDSL” file that exists on the disk. The current application has been developed in C++ language. It parses an already made “XDSL” file containing a Bayes Belief network and analyses it. It also load the data related to the initial samples from the Cohn-Kanade database. For each of the parameters it runs some tasks as shown: - determines the parent parameters of the current parameter - analyzes all the states of the parent parameters and determines the all possible combination of the states - for each state of the current parameter, passes through each of the possible combinations of the parents and: o create a query that includes the state of the current parameter o add the combination of parent parameters - 76 -
o call a specialized routine for computing the probability of existence of current parameter being in current state, given the parent parameters being in the states specified through the current combination o fills in the data in the Conditional Probability Table (CPT) of the current parameter - saves the data on the disk, in a file having the same name as the input file. The difference is that the saved model has the data in parameter CPTs. In order to be able to query the program for obtaining the probability associated to a certain state of the parameters, a C++ class tackles the data within the chosen parametric model (Listing 9). //--------------------------------------------------- 1: class model 2: { 3: public: 4: file a[1000]; 5: int n; 6: int NP; //no.of parameters 7: int NC; //no.of classes 8: public: 9: model():n(0){} 10: bool readDatabase(char*); 11: void show(void); 12: float computeP(cere&); 13: float computeP2(cere2&); 14: float computeP3(int,char*); 15: float computeP4(int,char*,bool); 16: int countExpression(char*); 17: bool isIncluded(FACS&,FACS&); 18: void select(query&,model&); 19: void emptyDatabase(void){n=0;} 20: int countFields(void){return n;} 21: int getClass(int,int); 22: void save(char*); 23: char* getFieldExpression(int); 24: private: 25: float compare(FACS&,FACS&); 26: void insertField(int,int,FACS&,char*); 27: }; //--------------------------------------------------- Listing 9. C++ class to handle the model data The class is used to manage all the processing on the data and to store the data related to the sample data from the initial sample database. An instance of the structure “file” stores internally the Action Units sequence, the description of the facial emotion and the value of the parameters. The definition is given in Listing 10. - 77 -
Page 1 and 2:
Automatic recognition of facial exp
Page 3 and 4:
Man-Machine Interaction Group Facul
Page 5 and 6:
Acknowledgements The author would l
Page 7 and 8:
Eye Detection Module ..............
Page 9:
List of tables Table 1. The used se
Page 12 and 13:
data taken from the Cohn-Kanade AU-
Page 14 and 15:
- The discussions on the current ap
Page 16 and 17:
ecognition in static pictures, for
Page 18 and 19:
In [Wang and Tang, 2003] the author
Page 20 and 21:
Data preparation Starting from the
Page 22 and 23:
Figure 2. Facial characteristic poi
Page 24 and 25:
The only additional time is that of
Page 26 and 27: African-American and three percent
Page 28 and 29: Table 4. The emotion projections of
Page 30 and 31: contains a large sample of varying
Page 32 and 33: Bayesian networks were designed to
Page 34 and 35: - correctly identify the goals of m
Page 36 and 37: In the final step of constructing a
Page 38 and 39: - renormalize the w ijk to assure t
Page 40 and 41: Principal Component Analysis The ne
Page 42 and 43: The term σ ij is the covariance be
Page 44 and 45: T The term rank( X ∗ X ) is gener
Page 46 and 47: numeric information. Usually, a neu
Page 48 and 49: defined as: ∆w = −η ∇ ji ji
Page 50 and 51: ∂E ∂a j = i E ∂a pk ∂a pk
Page 52 and 53: Spatial Filtering The spatial filte
Page 54 and 55: way high pass filter were used for
Page 56 and 57: module includes some routines for d
Page 59 and 60: IMPLEMENTATION Facial Feature Datab
Page 61 and 62: SMILE resides in a dynamic link lib
Page 63 and 64: FCP Management Application The Cohn
Page 65 and 66: Figure 13. Head rotation in the ima
Page 67 and 68: Table 6. The set of rules for the u
Page 69 and 70: [image width] [image height] ---- A
Page 71 and 72: Figure 17. The facial areas involve
Page 73 and 74: The functionality of the tool was b
Page 75: A small part of the output text fil
Page 79 and 80: There is another kind of structure
Page 81 and 82: performing classification of facial
Page 83 and 84: Figure 22. Sobel edge detector appl
Page 85 and 86: almost closed it obviously does not
Page 87 and 88: Figure 28. FCP detection The effici
Page 89 and 90: TESTING AND RESULTS The following s
Page 91 and 92: BBN experiment 2 “Detection of fa
Page 93 and 94: General recognition rate is 63.77%
Page 95 and 96: Recognition results. Confusion Matr
Page 97 and 98: 5 states model General recognition
Page 99 and 100: LVQ experiment “LVQ based facial
Page 101 and 102: ANN experiment Back Propagation Neu
Page 103 and 104: PCA experiment “Principal Compone
Page 105 and 106: Eigenvalues: Eigenvectors: Factor l
Page 107 and 108: Squared cosines of the variables: C
Page 109 and 110: - 109 -
Page 111 and 112: CONCLUSION The human face has attra
Page 113 and 114: REFERENCES Almageed, W. A., M. S. F
Page 115 and 116: Essa, A. Pentland, ‘Coding, analy
Page 117: Samal, A., P. Iyengar, ‘Automatic
Page 120 and 121: 61: } 62: } 63: float model::comput
Page 122 and 123: 119: for(k=0;k
Page 125 and 126: APPENDIX B Datcu D., Rothkrantz L.J
Page 127 and 128:
facial feature and store the inform
Page 129 and 130:
available as part of the knowledge
Page 131 and 132:
detailed in table 4. The dependency
Page 133:
[11] M. Turk, A. Pentland ‘Face r
show all

Recognition of facial expressions - Knowledge Based Systems ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?