Browsing by Author "Priyadarshini, Ishaani"
Now showing 1 - 3 of 3
Results Per Page
Sort Options
Item AI Cannot Understand Memes: Experiments with OCR and Facial Emotions(Computers, Materials & Continua, 2021-05-11) Priyadarshini, Ishaani; Cotton, ChaseThe increasing capabilities of Artificial Intelligence (AI), has led researchers and visionaries to think in the direction of machines outperforming humans by gaining intelligence equal to or greater than humans, which may not always have a positive impact on the society. AI gone rogue, and Technological Singularity are major concerns in academia as well as the industry. It is necessary to identify the limitations of machines and analyze their incompetence, which could draw a line between human and machine intelligence. Internet memes are an amalgam of pictures, videos, underlying messages, ideas, sentiments, humor, and experiences, hence the way an internet meme is perceived by a human may not be entirely how a machine comprehends it. In this paper, we present experimental evidence on how comprehending Internet Memes is a challenge for AI. We use a combination of Optical Character Recognition techniques like Tesseract, Pixel Link, and East Detector to extract text from the memes, and machine learning algorithms like Convolutional Neural Networks (CNN), Region-based Convolutional Neural Networks (RCNN), and Transfer Learning with pre-trained denseNet for assessing the textual and facial emotions combined. We evaluate the performance using Sensitivity and Specificity. Our results show that comprehending memes is indeed a challenging task, and hence a major limitation of AI. This research would be of utmost interest to researchers working in the areas of Artificial General Intelligence and Technological Singularity.Item Exploring the differences between human and machine intelligence: towards technological singularity(University of Delaware, 2021) Priyadarshini, IshaaniMany recent advancements in automation and other technologies can now be attributed to the use of Artificial Intelligence (AI), which has recently experienced significant advancements in capability and performance. ☐ Artificial Intelligence and related Machine Learning (ML) technologies are constantly evolving. While these advancements and their increasing incorporation into our daily lives are certainly to be marveled at, the use of this technology, like all technology advancements has both positive and negative effects. Creating accurate image classifiers has significant value for the medical researcher trying to differentiate benign from malignant cells, but other uses can have horrific outcomes such as in the case of ill-conceived aircraft flight control systems or the reduction in privacy brought about by facial recognition systems. ☐ The roots of our research began as an exploration of the use of AI in the cybersecurity domain specifically aimed at identifying a viable fourth factor for Multi-factor Authentication (MFA). What started as something you might call a "reverse Turing test" became an examination of about a dozen things that a human might be able to do versus what a machine might be able to do. This was followed by several focused investigations of the use of this technology (called cyberpsychology) to enhance conventional cybersecurity use cases (e.g. authentication). ☐ These insights into the currently perceived differences between human and artificial intelligence led to a second line of research to better explore more futuristic possibilities. Possibilities, not those as seen from a science fiction standpoint, but those we can use to better understand what hard science contributions have been or need to be made to better understand these differences. Key foundational studies are made in these areas to better define the problem area itself, the nature of the hard science contributions, and how machine intelligence can be classified and measured. And last, the need to better understanding where the ever-changing boundary is between human and artificial intelligence, and if the location of that boundary can be estimated over time.Item Features and architecture of the modern cyber range: a qualitative analysis and survey(University of Delaware, 2018) Priyadarshini, IshaaniCybersecurity is one of the prominent global challenges due to significant increase in the number of cyber-attacks over the last few decades. To protect enterprises, personal data, productivity and to ensure a safe environment for work, cybersecurity awareness is very important. To prevent cyber infection, ad wares and to provide a consolidated solution, cybersecurity training is vital. Cybersecurity awareness and cyber security training are promoted by hyper-realistic virtual environments termed as cyber ranges. This study highlights the concept of cyber ranges, mainly in two sections. In the first section, several cyber ranges have been taken into consideration to identify significant attributes and classification based on several properties. An Ideal Cyber Range has been proposed taking into account the important parameters a cyber range should incorporate. In the second section, the Cyber Range at the University of Delaware has been introduced along with its fundamental working. The extensive study takes into account the components, scenarios and capabilities of the newly found cyber range. Finally, a comparison has been carried out between the already proposed Ideal Cyber Range and the Cyber Range at the University of Delaware to comprehend the proximity between the two. This would be of great assistance in future for determining the strength and weaknesses of the Cyber Range at University of Delaware considering what the range is capable of at present and what could be done in future for the Range to become Ideal. We rely on different graphs to bring out the variation between the Ideal Cyber Range and The Cyber Range at the University of Delaware on basis of certain parameters.