Rapid Stereo-Vision Enchanced Face Recognition

[Creator]

This is my third schientific article in english. It was submitted to the International Conference on Image Processing (ICIP), which will take place in Hong-Kong this year.

ICIP_10.pdf

(857.37 KiB) Downloaded 24 times

P.S.

Please see the suplementing video

[Creator]

The ICIP 2010 Technical Program Committee has completed the review process and we are pleased to inform you that the manuscript listed above has been ACCEPTED for presentation at the conference. Congratulations!

[Creator]

Importance/Relevance: Of sufficient interest

Comment on Importance/Relevance: This paper describes an approach for face recognition based on PCA. This approach uses both intensity and disparity images.

Novelty/Originality: Moderately original

Comment on Novelty/Originality: The original contribution of this approach results from the combination in real-time of disparity and intensity images.

Technical Correctness: Probably correct

Experimental Validation: Limited but convincing

Clarity of Presentation: Very clear

Reference to Prior Work: References adequate

General Comments to Authors: This paper is very clear, and can be read easily. The main issue with the approach described is its scalability. Results are presented with a database of only 34 faces. The video shows the detection and recognition of more than one face. However the paper does not address that issue and in particular the problem of face segmentation. Those elements would certainly add to the technical value of the paper.

[Creator]

Importance/Relevance: Of broad interest

Novelty/Originality: Moderately original

Technical Correctness: Probably correct

Comment on Technical Correctness: Please look into your defined functional in Section 3. It appears to be a typo there. In Section 5, is the Gaussian noise white? If so, please indicate.

Experimental Validation: Limited but convincing

Clarity of Presentation: Clear enough

Reference to Prior Work: References adequate

[Creator]

Importance/Relevance: Of broad interest

Novelty/Originality: Moderately original

Comment on Novelty/Originality: Although this paper presents a combination of knowing techniques it has some original assumptions.

Technical Correctness: Probably correct

Experimental Validation: Lacking in some respect

Comment on Experimental Validation: The authors used a limited size database which was created by themselves. They should provide extended experimental results in order to completely prove their claim.

Clarity of Presentation: Very clear

Reference to Prior Work: References adequate

[Creator]

The corrected final version:

ICIP_2010.pdf

(901.53 KiB) Downloaded 7 times