Category: capstone2022

 
All Too Human is a project that investigates human perception in the visual arts, particularly the variations in perception between machine-generated and human-created images. As GAN-based image synthesis advances, our understanding of visual perception becomes increasingly multifaceted. This leads to debates in the field of art around notions such as authorship, vision, intelligence, and–most importantly–creativity, which has been regarded as one of the distinctive human capabilities.
In a similar vein, several versions of the “Turing test” have been proposed to examine if a machine can demonstrate such human capabilities. The Bot or Not project, for instance, tests whether users can distinguish between AI-written and human-written poems. All Too Human poses a similar question, but with paintings instead of poems. Admittedly, seeing an image is drastically different from reading a text. We cannot identify the difference between human-written and computer-written text from its appearance alone. As a result, what matters in the Turing test for poetry is primarily semantics and the usage of syntax. The Turing test for paintings, however, is a far more complex subject that cannot be boiled down to one single factor. Hence, this project aims to probe which types of AI-generated images appear to be more human-made, and what factors have influenced such visual perceptions.
This web-based project displays AI-generated paintings and human-made paintings in a random sequence, allowing the viewers to guess whether the image was created by an algorithm or a human. The human image set encompasses a wide range of styles (Realism, Impressionism, Surrealism, etc.) and genres (portrait, abstract, landscape), whereas the machine image set includes results of different algorithms and training datasets. I primarily used two image-generating models–StyleGAN2 and CLIP+VQGAN. For StyleGAN2, I utilized a WikiArt dataset consisting of 1) 40,000 paintings that include a comprehensive anthology of art history and 2) the selection of early twentieth-century abstract paintings. On the other hand, CLIP is trained with 400 million image-text pairs that contain diverse images such as photos, paintings, and even screenshots. Since the project presents images without any external context, say, their title or the artist’s name, the viewers have to intuitively guess the authorship of these images. To collect more genuine responses, the website does not reveal the creator of each image till the very end, nor does it inquire about what factors influenced the viewer’s judgment. A separate leaderboard section lists the creator of each painting and the proportion of user responses.
In summation, the focus of this project is less on how adept algorithms have become at creating plausible images; instead, this experiment seeks to uncover some insights into common perceptions and expectations about AI in the context of visual arts. What does it mean if an AI-generated image passes this test? What does it mean for humans to have a visual culture in which we cannot tell whether the images were created by humans or machines?

 
RAMify is an interactive VR storytelling experience that revolves around futuristic technology for memory editing and removal. From written language to photography, human culture has long relied on external tools for storing personal and societal memory. With the ascent of digital culture and its capacity for seamless archiving, not only has the effectiveness of these tools increased but, as some scholars argue, the human behavioral default has shifted from forgetting to remembering. RAMify explores the possible consequences of embracing future advanced memory technologies.

The story of RAMify takes place in a cyberpunk society, where memory implants and memory retrieval technologies have become commonplace. The player plays as a memory architect working for the MindGen Corporation, currently tasked with fulfilling a dying woman’s last wish of recovering the memory of her long-lost son. In the course of gameplay, the player must face complex ethical questions about whether to preserve, alter or delete the memories, as they uncover the relationship between the mother and the son. RAMify’s narrative immersion is delivered through a combination of the Passthrough AR feature and VR to emphasize the likeliness of such scenarios occurring and create a higher level of immersion for the experience.

Though set within the convention of the futuristic cyberpunk genre, RAMify addresses already existing dilemmas around technology and memory. Within the interactive experience, the ability to navigate through an individual’s memory represents the possible future of memory externalization. Exploring both the positive and negative ramifications of hypothetical memory tech, it rejuvenates the concern of the cyberpunk genre in a distinctly contemporary setting, highlighting the choices we make when using our technologies in order to remember or forget.

Through showing the possible consequences that memory over-externalization would bring to individuals and society in the future era, my project would raise people’s awareness on remembering to forget, and therefore avoid the dystopian cyberpunk scenario that the lack of forgetting would forge.

 
In this video-based project, the examination of celebrity-influence on the general public as consumers is examined and taken into account on a psychological and subconscious level. Kanye West, as well as other celebrities are used as common examples of just how often celebrities are partnered with certain brands, both parties profiting off of their reputation and using fame and relevance to capitalize off of consumers. This takes away the individual choice of the consumer to make their own choices based on their own genuine preferences, and hinders their chances of going out and finding smaller shops that have more personal roots and a deeper connection with their customers. The videos for this project have been posted on an anonymous TikTok account. For the purposes of the exhibition, the videos have been compiled into one file on repeat, demonstrating the cycle of consumers following in the footsteps of what celebrities tell them to purchase and support. The four videos hope to take the viewer through a journey of seeing the issue, recognizing the problem, understanding one of the many causes of the problem, and coming to terms with a solution. Taking the viewer through a visual journey is a way for them to look at the situation from a third person perspective, since the viewers are usually encountering this phenomenon in the first person, unable to see how the situation looks from afar. By viewing this cycle from a visual third person perspective, they are hopefully able to recognize the issue with their own eyes and hopefully feel the desire to make a change in their consumption habits. In accordance with this change in consumption habits, during my project, I created a spreadsheet of local shops in any area that anyone could collaborate in, and in the span of a week, about 12 random people added to the list, listing smaller shops that may not get as much attention in comparison to larger brands. This then also gives these small local shops the chance to gain new customers that come from viewing the spreadsheet, executing impact beyond an academic project. The spreadsheet has more opportunity to grow during the exhibition, as viewers within the exhibition watch all four videos and are able to participate. It should be normalized for people to go out and find unique stores within their cities and spread the word, instead of putting in minimal effort and follow the wave of relevant celebrities. Celebrities endorsing and advertising for certain brands are not in the interest of trying to show consumers the best product on the market, but instead to work in their best interest by furthering their financial standing by partnering with large brands, leaving smaller businesses little room to compete.

 
See and be Seen is a project that investigates the echo chamber phenomenon **in** the context of the Chinese social media environment. It focuses on how the public sphere of social media, such as “Weibo trending,” enables users to create online social influence with their posts, emotions and personal views. The weighted evaluation of Weibo trending, based on user interaction with posts, can easily cause view polarization. The ranking of posts and comments against a hashtag is decided by the calculation of interaction factors, which leads to the phenomenon that the most obvious messages and posts either share similar attitudes, or that the most extreme arguments cause controversy. This affects media ecology and the public opinion environment, including misinformation, cyberbullying, group polarization, etc. However, the interaction with posts is only reflected as numbers on Weibo. Invisible algorithmic rankings dictate which posts attract attention. Because of that, it is difficult for users to understand how their interaction with a post is increasing its exposure and popularity. Composed of a mobile-end interaction and a visualization website, *See and be Seen* aims for a creative perspective through which we can rethink the influence of Weibo’s interaction mechanism design on the visibility of each comment.

The interaction taking place on the mobile phone can trigger the animation on the webpage so that the connection between mobile-end and visualization webpage can make the influence more explicit and intuitive for the audience. It visualizes the influence of time, likes, and replies on the exposure of each comment under the trending hashtag. By scanning the QR code on the webpage, the audience can get access to the mobile-end part, which is a list of comments under the example hashtag. As the audience posts a new comment, there will appear a new house on the webpage. The earlier the comment, the more centered its position will be. The size and level of each house depends on likes and replies it receives.

When we use Weibo in our daily lives, we always take it for granted that some comments are more visible in the front row because they get more likes and comments, ignoring the decisive role of the algorithm. However, such a mechanism can induce the clickbait issues driven by the attention economy. It is common to see some accounts deliberately give controversial comments to stir emotions and attacks to earn more views, and there are many replies under this comment, making it more visible to others. If people learn about how the Weibo algorithm decides the exposure of each comment based on interaction index, they might reevaluate those comments with high ranks, and think twice before giving likes or replies. Instead of accepting the default order set by Weibo, which is sorting comments by popularity, users may see comments in reverse order to see a variety of different comments.

 
This project utilizes the technology of Virtual Reality to materialize and introduce the different interpretations of the Chinese concept “jianghu” (江湖). It is inspired by the transposing representations and multiple dimensions of interpretations of the text “jianghu ”. It is one word that carries much philosophical thinking and describes social conditions. The interpretation of this word originates from the Daoist discussion by Zhuangzi of spiritual liberation and individual freedom. It later spans the genre of Chinese martial arts (Wuxia) novels depicting the netherworld of misfits who rejected systems of authority. In modern times, it is used to describe the paradoxical side of relationships within societies. It portrays how different perspectives can lead to different interpretations, and how different dimensions of thinking can lead to various media representations. These features of Jianghu align with many characteristics of the VR technology that allows for connecting different dimensions of worlds and different perspectives together. So far, many projects have focused on one aspect among the multiple transposing interpretations of jianghu. But little attention has been put to the introduction about the connections between the philosophical, ethnical, and aesthetic aspects of these different representations. Thus, in this project, I utilize Virtual Reality to materialize and introduce the different dimensions of the concept “jianghu” (江湖).

The player became the Taoism philosopher Zhuangzi in this project. He dreamed that he became a fish. But he did not know that he became a fish at first. In the exploration of the different scenes of “jianghu” he encountered, he gradually realized where he is, what is jianghu, and what he has become. Eventually, he went back to the river and lake and finally became ready to forget about all the attachments in the jianghu and achieve more understanding of spiritual liberation.