1 |
Robust recognition of facial expressions on noise degraded facial imagesSheikh, Munaf January 2011 (has links)
<p>We investigate the use of noise degraded facial images in the application of facial expression recognition. In particular, we trained Gabor+SVMclassifiers to recognize facial expressions images with various types of noise. We applied Gaussian noise, Poisson noise, varying levels of salt and pepper noise, and speckle noise to noiseless facial images. Classifiers were trained with images without noise and then tested on the images with noise. Next, the classifiers were trained using images with noise, and then on tested both images that had noise, and images that were noiseless. Finally, classifiers were tested on images while increasing the levels of salt and pepper in the test set. Our results reflected distinct degradation of recognition accuracy. We also discovered that certain types of noise, particularly Gaussian and Poisson noise, boost recognition rates to levels greater than would be achieved by normal, noiseless images. We attribute this effect to the Gaussian envelope component of Gabor filters being sympathetic to Gaussian-like noise, which is similar in variance to that of the Gabor filters. Finally, using linear regression, we mapped a mathematical model to this degradation and used it to suggest how recognition rates would degrade further should more noise be added to the images.</p>
|
2 |
Robust recognition of facial expressions on noise degraded facial imagesSheikh, Munaf January 2011 (has links)
<p>We investigate the use of noise degraded facial images in the application of facial expression recognition. In particular, we trained Gabor+SVMclassifiers to recognize facial expressions images with various types of noise. We applied Gaussian noise, Poisson noise, varying levels of salt and pepper noise, and speckle noise to noiseless facial images. Classifiers were trained with images without noise and then tested on the images with noise. Next, the classifiers were trained using images with noise, and then on tested both images that had noise, and images that were noiseless. Finally, classifiers were tested on images while increasing the levels of salt and pepper in the test set. Our results reflected distinct degradation of recognition accuracy. We also discovered that certain types of noise, particularly Gaussian and Poisson noise, boost recognition rates to levels greater than would be achieved by normal, noiseless images. We attribute this effect to the Gaussian envelope component of Gabor filters being sympathetic to Gaussian-like noise, which is similar in variance to that of the Gabor filters. Finally, using linear regression, we mapped a mathematical model to this degradation and used it to suggest how recognition rates would degrade further should more noise be added to the images.</p>
|
3 |
Výchovně vzdělávací program o civilní ochraně pro žáky základních škol Slovenské republiky / Pedagogic & educational programme of civil defence of population for elementary school pupils in the Slovak republic.UŠIAKOVÁ, Michaela January 2013 (has links)
Theme of our thesis is in generally not much mapped issues, regarding education and prepares people and especially pupils of elementary schools in Slovak Republic. Theme I chose because of my long experience with issues relating to the protection of human health and life associated with the provision of the first aid and solutions incidents as medical paramedic, voluntary fire brigade and civil protection units. The aim of this thesis is to identify and assess expert opinion on civil protection and crisis management of education and teachers' preparedness to deal with emergencies and the need to introduce a separate subject civil protection in elementary schools Slovak Republic. Another aim is to analyze teachers interested in establishing a separate school subject civil protection in elementary schools Slovak Republic. Based on research in the practical part we set other goals: create a draft manual educationally education program on civil population protection for upper elementary schools and its subsequent implementation in the selected elementary school in Dojč. A handbook could serve for the preparation of pupils for self-defense and mutual assistance in the development of emergencies. Research was realized using qualitative and quantitative, analytical and synthetic, inductive-deductive methods, in the form of semi-structured interviews with experts in civil protection of the population and statistical survey among teachers in elementary schools in Slovak Republic. Researches question ?It is considered important to establish a separate subject of civilian protection in elementary schools of Slovak Republic?? Was it answered and statistical verificated. The result of statistical investigation was positive; this means that teachers are agreed on the necessity of establishment an independent subject of civil protection. On-backed talks, we have concluded that the instruction relating to the protection of civilian populations is inadequate. Teachers haven't got any sources to draw from and also their knowledge are not satisfactory. We believe that the practical ignorance and inaction after the unusual event in its first phase could have fatal consequences.
|
4 |
Robust recognition of facial expressions on noise degraded facial imagesSheikh, Munaf January 2011 (has links)
Magister Scientiae - MSc / We investigate the use of noise degraded facial images in the application of facial expression recognition. In particular, we trained Gabor+SVMclassifiers to recognize facial expressions images with various types of noise. We applied Gaussian noise, Poisson noise, varying levels of salt and pepper noise, and speckle noise to noiseless facial images. Classifiers were trained with images without noise and then tested on the images with noise. Next, the classifiers were trained using images with noise, and then on tested both images that had noise, and images that were noiseless. Finally, classifiers were tested on images while increasing the levels of salt and pepper in the test set. Our results reflected distinct degradation of recognition accuracy. We also discovered that certain types of noise, particularly Gaussian and Poisson noise, boost recognition rates to levels greater than would be achieved by normal, noiseless images. We attribute this effect to the Gaussian envelope component of Gabor filters being sympathetic to Gaussian-like noise, which is similar in variance to that of the Gabor filters. Finally, using linear regression, we mapped a mathematical model to this degradation and used it to suggest how recognition rates would degrade further should more noise be added to the images. / South Africa
|
Page generated in 0.1802 seconds