With the recent advances in computer vision and pattern recognition, methods from these fields are successfully applied to solve problems in various domains, including health care and social sciences. In this thesis, two such problems, from different domains, are discussed. First, an application of computer vision and broader pattern recognition in physical therapy is presented. Home-based physical therapy is an essential part of the recovery process in which the patient is prescribed specific exercises in order to improve symptoms and daily functioning of the body. However, poor adherence to the prescribed exercises is a common problem. In our work, we explore methods for improving home-based physical therapy experience. We begin by proposing DyAd, a dynamically difficulty adjustment system which captures the trajectory of the hand movement, evaluates the user's performance quantitatively and adjusts the difficulty level for the next trial of the exercise based on the performance measurements. Next, we introduce ExerciseCheck, a remote monitoring and evaluation platform for home-based physical therapy. ExerciseCheck is capable of capturing exercise information, evaluating the performance, providing therapeutic feedback to the patient and the therapist, checking the progress of the user over the course of the physical therapy, and supporting the patient throughout this period. In our experiments, Parkinson patients have tested our system at a clinic and in their homes during their physical therapy period. Our results suggests that ExerciseCheck is a user-friendly application and can assist patients by providing motivation, and guidance to ensure correct execution of the required exercises. As the second application, and within computer vision paradigm, we focus on visual complexity, an image attribute that humans can subjectively evaluate based on the level of details in the image. Visual complexity has been studied in psychophysics, cognitive science, and, more recently, computer vision, for the purposes of product design, web design, advertising, etc. We first introduce a diverse visual complexity dataset which compromises of seven image categories. We collect the ground-truth scores by comparing the pairwise relationship of images and then convert the pairwise scores to absolute scores using mathematical methods. Furthermore, we propose a method to measure the visual complexity that uses unsupervised information extraction from intermediate convolutional layers of deep neural networks. We derive an activation energy metric that combines convolutional layer activations to quantify visual complexity. The high correlations between ground-truth labels and computed energy scores in our experiments show superiority of our method compared to the previous works. Finally, as an example of the relationship between visual complexity and other image attributes, we demonstrate that, within the context of a category, visually more complex images are more memorable to human observers.

Due to its simplicity and deployability, password authentication is today's most common way of authentication. In conjunction with increasing numbers of accounts per user, the amount of passwords to be remembered is rising as well. This puts a noticeable strain on human memory which users attempt to mitigate by writing them down, reusing them or selecting overly simple ones. In order to prevent such behavior and the security issues it is accompanied by, finding ways to generate memorable passwords is imperative. The conducted systematic literature review aimed to identify to which extent different password creation strategies are facilitating the generation of memorable passwords. Several search term combinations were used to probe four scientific databases for peer-reviewed articles that satisfy distinct selection criteria. Afterwards, backward snowballing was conducted and references of already accepted publications were checked against identical selection criteria. Eventually, 61 accepted articles underwent a qualitative data analysis by means of grounded theory. The analysis showed that different composition strategies entailed substantial differences in memorability. Those that infused passwords with deeper meaning to the user were found easy to remember whereas failing to infuse meaning impeded recall. Overall, user-generated passwords turned out be more memorable than system-generated ones. / På grund av sin enkelhet och användbarhet är lösenordsautentisering dagens mest använda sätt att autentisera sig. Antalet lösenord som ska minnas ökar i samband med ett stigande antal konton per användare. Detta utgör en tydlig belastning för det mänskliga minnet vilket användarna försöker lindra genom att skriva ner lösenord, återanvända dem eller välja alltför enkla varianter. Att hitta sätt att skapa lösenord som är lätta att komma ihåg är en nödvändighet för att förebygga liknande beteenden samt de säkerhetsproblem som de är förknippade med. Den utförda systematiska litteraturstudien hade som mål att identifiera i vilken omfattning olika strategier för lösenordskonstruktion underlättar skapandet av minnesvärda lösenord. Ett flertal kombinationer av sökbegrepp användes för att söka i fyra vetenskapliga databaser efter granskade artiklar som uppfyller tydliga urvalskriterier. Efteråt användes backward snowballing och referenser av redan accepterade publikationer kontrollerades med samma urvalskriterier. Till slut genomgick 61 accepterade artiklar en kvalitativ dataanalys med hjälp av grounded theory. Analysen visade att olika kompositionsstrategier förde med sig betydande skillnader gällande hur minnesvärda resulterande lösenord var. Strategier som fyllde lösenord med djupare betydelse för användaren ansågs vara lätta att minnas medan det var svårare att komma ihåg dem när ingen betydelse ingjutits. Över lag visade sig användargenererade lösenord vara lättare att komma ihåg än systemgenererade.

Interactive fisheye views use distortion to show both local detail and global context in the same display space. Although fisheyes allow the presentation and inspection of large data sets, the distortion effects can cause problems for users. One such problem is lack of memorability the ability to find and go back to objects and features in the data. This thesis examines the possibility of improving the memorability of fisheye views by adding historical information to the visualization. The historical information is added visually through visit wear, an extension of the concepts of edit wear and read wear. This will answer the question Where have I been? through visual instead of cognitive processing by overlaying new visual information on the data to indicate a users recent interaction history. This thesis describes general principles of visibility in a space that is distorted by a fisheye lens and defines some parameters of the design space of visit wear. Finally, a test system that applied the principles was evaluated, and showed that adding visit wear to a fisheye system improved the memorability of the information space.

Interactive fisheye views use distortion to show both local detail and global context in the same display space. Although fisheyes allow the presentation and inspection of large data sets, the distortion effects can cause problems for users. One such problem is lack of memorability the ability to find and go back to objects and features in the data. This thesis examines the possibility of improving the memorability of fisheye views by adding historical information to the visualization. The historical information is added visually through visit wear, an extension of the concepts of edit wear and read wear. This will answer the question Where have I been? through visual instead of cognitive processing by overlaying new visual information on the data to indicate a users recent interaction history. This thesis describes general principles of visibility in a space that is distorted by a fisheye lens and defines some parameters of the design space of visit wear. Finally, a test system that applied the principles was evaluated, and showed that adding visit wear to a fisheye system improved the memorability of the information space.

The society today is dependent on information technology and with the help of the technologymakes it easier to access information. Due to the constantly growing network environment,various techniques of accessing and handling information have developed. One of the mostused solutions to access and protect information is by using a password. The purpose of apassword is to protect sensitive and important data from unauthorized users who intentionallyor accidentally access the system. This can lead to unsolicited modifications of the original dataas well as unauthorized access of confidential information. Humans are those who design theinformation security, but at the same time the ones who are the weakest link in the securitychain. To prevent unauthorized access it is important to have a strong and tamper proofpassword. A good password should be easy to remember, hard to guess by others and bedifficult to predict by a person or software. The goal in this study is to find a good balancebetween a memorable and a secured password. The study will compare three types ofconstructions for password, own set, modified dictionary and association against each other tofind the one which is the strongest and the most memorable.

In an era where digital security concerns are paramount, understanding users' behaviors and attitudes towards password management is crucial. This research examines users' perception of their password habits in terms of security, memorability, and usability. The study encompasses a comprehensive survey and interviews conducted with individuals across various professional backgrounds and age groups. A total of 87 online users responded to the questionnaire and ten individuals were interviewed to gather more in-depth data. The study aimed to understand how users perceive the trade-offs between security, memorability, and usability in their password habits and what factors influence their choices in creating and managing passwords. The findings reveal nuanced insights into users' password practices. While users acknowledge the importance of security, they often prioritize memorability and usability over stringent security measures. Factors such as convenience, familiarity, and personal preferences significantly influence password creation and management. Despite being aware of security risks associated with weak passwords and password reuse, users commonly engage in these practices due to the challenges posed by complex password requirements and the sheer volume of passwords needed for various accounts. Overall, this research underscores the importance of understanding users' perspectives on password habits to develop more effective strategies for promoting password security. By bridging the gap between user behavior and security policies, organizations may be able to tailor interventions that align with user preferences, thereby fostering a more secure online environment.

The ability to memorize and recognize edible prey from inedible prey is essential for an individuals survival. Many species use aposematic signals for their defense. These are most often represented by distinctive colors or contrast patterns. The aim of this study was to find out in which time the memory consolidation for aposematic pray is achieved and if color or pattern increase its memorability. The choosen model species was the Great tit (Parus major). The comparision was done between handreared naïve birds and wild-caught adults of different age and sex. During the discriminatory task of consolidation experiment, birds were simultaneously presented with palatable and unpalatable prey in the form of paper dummies of bugs, differing in color (red versus green). The birds were divided into three groups with a different interval (0, 1 or 3 hours) for consolidation. The results of consolidation test show that adult birds were more successfull in solving the task than juvenile birds. The only difference between the experimental groups was that the group with one-hour interval achieved better results than other groups. The effect of color of palatable and unpalatable prey on discrimination learning was also found only in adults. Memorability of warning signals was tested using paper bugs of...

Social Networking Sites have taken the world by storm with an aggressively growing market. People are spending more and more time with ever more alluring algorithms to keep a high interest for increased profit margins and revenue. However, this brings forth societal issues regarding attention span. Herbert. A. Simons (1971) theorized that as people gain information, it is in exchange for attention in an attention economy.With these uprising issues, we set out to discover if there were ways to alleviate the attention required to understand the content. In design, minimalism as a design concept helps to solve these issues by communicating as effectively as possible and skipping "the small talk." Further, the massive growth SNS has experienced in the 2010s leaves a market opportunity to streamline advertisements to communicate with its audience more effectively.The research set out to answer two questions: “How do end-users perceive minimalistic design principles to affect memorability in SNS advertisements?” and “How do end-users perceive minimalistic design principles to affect the attention required to understand an advertisement on SNSs?” We constructed interviews based on previous research and conducted eight interviews. The interviews were then thematically analyzed, where four significant themes occurred. "Positive Perception of Attention Required", "Positive Perception of Effects on Memorability", "Influential Individual Content Creators" and “Effects on Minimalism by Context”. The findings revealed a positive perception of how minimalistic design principles affected the attention required to understand the content. Likewise, it was found that people had similar perceptions of the effects on memorability. Further, it was revealed that a majority believe individual content creators can alleviate societal issues regarding attention.

This thesis investigates if a memorability-influenced design could improve the effectivity of a user when navigating through a pharmacy webshop. This was measured by calculating the error ratio, completion time, and number of clicks when solving a task. To arrive at a definitive conclusion within the context of this thesis, one prototype with a memorability-influenced design was created, and another prototype with a design influenced by an already existing pharmacy webshop. The design that incorporated memorability was informed by an exploration of mental models and other design attributes aimed at augmenting memorability. These prototypes underwent evaluation through an A/B testing methodology. A total of ten participants, equally divided into two groups of five, were involved in this A/B test. All participants were adults aged over 18, possessing prior experience in procuring products and prescriptions from a physical pharmacy store. The results after analyzing the data from the A/B testing shows no significant difference between the prototypes when it comes to effectivity, which leaves the result inconclusive. Hence the appliance of memorability-influenced design traits did not lead to any difference in the A/B test. Even though this study could not find any correlation between memorability enhanced design and user’s effectivity, we found other aspects of the result that were interesting and helpful when it comes to user experience. From the survey that each person took after finishing the A/B test, it was discovered that prototype A, the one with memorability design traits, was experienced more positively than the other prototype without the amplified memorability design traits. The users described prototype A as easy and smooth while prototype B was described as cluttered and confusing. The results from the survey conclude that prototype A, which was designed with memorability design traits, enhanced the user experience in a positive way. / Detta examensarbete undersökte hur memorabilititets influerad design kan förbättra en användares effektivitet när denna navigerar sig genom ett gränssnitt för apotekweb- shoppar. Detta undersöktes genom att beräkna error ratio, sluttid och antal klick för att lösa en uppgift. För att komma fram till en definitiv slutsats inom ramen för detta examensarbete, skapades en prototyp med memorabilitets influerad design samt en prototyp med en design influerad från ett redan existerande gränssnitt för en apotekwebbshop. Prototy- pen som influerades av memorabilitet skapades genom att utforska mentala modeller och andra designattribut som kan öka memorabiliteten. Dessa prototyper utvärderades genom en A/B testning metod. Totalt deltog 10 stycken deltagare i A/B testningen. Dessa delades upp i två grupper om 5 deltagare per grupp. Varje grupp fick testa en varsin prototyp. Alla deltagare var över 18 år gamla och hade tidigare erfarenhet gällande att handla produkter och recept från ett fysiskt apotek. Efter att ha analyserat datan från A/B testningen visar resultaten på att det inte finns en signifikant skillnad mellan prototyperna när det kommer till effektivitet. Till- lämpningen av memorabilitets influerad designattribut härledde inte till några skillna- der mellan prototyperna i denna A/B testning. Även fast denna studie inte kunna hitta någon korrelation mellan memorabilitets influerad design och en användares effektivitet, hittades andra aspekter av resultatet som var intressanta och användbara när det kommer till användarupplevelsen. Från enkäten som varje person deltog i efter att ha genomfört A/B testningen, upptäcktes det att prototypen som designats med memorabilitets attribut (prototypn A) upplev- des mer positivt än den andra prototypen (prototyp B) utan memorabilitets attribut. Deltagarna beskrev prototyp A som enkel och smidig med prototyp B beskrevs som rö- rig och förvirrande. Resultatet från enkäten sammanfattar att prototyp A, prototypen med memorabilitets attribut, förbättrade användarupplevelsen på ett positivt sätt.

With the growth of digital information and technology, more websites require user authentication to access individuals' sensitive information. The main form of authentication are passwords which if chosen unwisely can easily be guessed or cracked by someone else. This is why it is important to create safe yet memorable passwords. The goal of this report was to develop a program that is able to detect commonly used patterns within passwords in order to transform a specified user inputted password into a more secure password without sacrificing the memorability of the original input. This was realized by analyzing lists of common passwords and partaking in a literature study within this field to identify the patterns present within those password lists. Based on the pattern analysis the program could be designed and developed using python to transform a password into three different password classes with differing levels of security and memorability. The password cracking software hashcat as well as online resources were used to estimate the time it would crack each set of passwords to then gain an understanding of the security levels between them. Results show that it is possible to perform password generation without sacrificing too much security while still having the passwords somewhat memorable. / Med framväxten av digital information och teknik kräver fler webbplatser användarautentisering för att få tillgång till individers känsliga information. Den huvudsakliga formen av autentisering är lösenord, och om de väljs oklokt, kan lätt gissas eller knäckas av någon annan. Det är därför det är viktigt att skapa säkra men ändå minnesvärda lösenord. Målet med denna rapport var att utveckla ett program som kan upptäcka vanliga mönster i lösenord för att omvandla ett specifikt användar inmatat lösenord till ett säkrare lösenord utan att offra minnesvärdheten av den ursprungliga inmatningen. Detta förverkligas genom att analysera listor med vanliga lösenord och delta i en litteraturstudie inom detta område för att identifiera mönstren som finns i dessa lösenordslistor. Baserat på mönsteranalysen kunde programmet designas och utvecklas med Python för att omvandla ett lösenord in till tre olika lösenordsklasser med olika nivåer av säkerhet och minnesvärdhet. Mjukvaran ”hashcat” (ett lösenordsåterställningsverktyg) samt resurser från nätet användes för att uppskatta den tid det skulle ta att knäcka varje uppsättning av lösenord för att sedan få en förståelse för säkerhetsnivåerna mellan dem. Resultaten visar att det är möjligt att digitalt generera lösenord utan att offra för mycket säkerhet, samtidigt som lösenorden behåller minnesvärdheten.