Investigating Robustness of DNNs - Revision history

Jens at 10:16, 16 October 2016

2016-10-16T10:16:11Z

Jens at 08:53, 17 December 2014

2014-12-17T08:53:33Z

Jens at 07:34, 17 December 2014

2014-12-17T07:34:50Z

Jens: Created page with "{{StudentProjectTemplate |Summary=This master thesis project aims at characterizing sensitivity to classification of images (based on deep neural networks). |Keywords=deep neu..."

2014-12-17T06:25:01Z

Created page with "{{StudentProjectTemplate |Summary=This master thesis project aims at characterizing sensitivity to classification of images (based on deep neural networks). |Keywords=deep neu..."

New page

{{StudentProjectTemplate
|Summary=This master thesis project aims at characterizing sensitivity to classification of images (based on deep neural networks).
|Keywords=deep neural networks, robustness
|TimeFrame=Spring 2015
|Prerequisites=Learning Systems, Data Mining
|Supervisor=Jens Lundström, Stefan Byttner,
|Level=Master
|Status=Open
}}

@@ Line 14: / Line 14: @@
 |Prerequisites=Learning Systems, Data Mining, Parallel programming
 |Supervisor=Jens Lundström, Stefan Byttner,
 |Level=Master
-|Status=Open
+|Status=Finished
 }}
 Deep Neural Networks (DNNs) have gained much interest during the last years. Among many sucessful applications, DNNs have shown outstanding performance in the task of learning feature representations and classification of images. A state-of-art, high accurate, neural network trained to classify 1.2 million images using 60 million parameters and 650,000 neurons was developed by Hinton et al. (2012). However, recent findings reveal delicate difficulties on noise robustness in DNNs, Szegedy et al (2013). The purpose of the thesis is to characterize the sensitivity of DNNs and potentially make suggestions on how robustness can be achieved.

@@ Line 17: / Line 17: @@
 |Status=Open
 }}
-Deep Neural Networks (DNNs) have gained much interest during the last years. Among many sucessful applications, DNNs have shown outstanding performance in the task of learning feature representations and classification of images. A state-of-art, high accurate, neural network trained to classify 1.2 million images using 60 million parameters and 650,000 neurons was developed by Hinton et al. (2012). However, recent findings reveal delicate difficulties on noise robustness in DNNs, Szegedy et al (2013). This master thesis project aims at two related studies. Firstly, the master student will investigate how meaningsless, to human, images are classified with high confidence using DNNs, as reported by other studies. Secondly, the student will investige DNNs misclassifications of images with small pertubations, not visible to humans. Moreover, the student is also encouraged to apply image preprocessing methods in order to increase classification accuracy.
+Deep Neural Networks (DNNs) have gained much interest during the last years. Among many sucessful applications, DNNs have shown outstanding performance in the task of learning feature representations and classification of images. A state-of-art, high accurate, neural network trained to classify 1.2 million images using 60 million parameters and 650,000 neurons was developed by Hinton et al. (2012). However, recent findings reveal delicate difficulties on noise robustness in DNNs, Szegedy et al (2013). The purpose of the thesis is to characterize the sensitivity of DNNs and potentially make suggestions on how robustness can be achieved.
 Four work packages are suggested: