Tag Archives: evaluation

Design Iterations – Evaluation

As a consequence of all my testing I have drawn some conclusions on the highlights and flaws of my installation and the overall design iterations unit. The aim of this unit was to create an understanding of the iterative design process and how it can be implemented when designing and creating a product or service.  The process includes; requirements gathering, analysis, design and prototyping, testing and evaluating. I believe the design iteration process helped me create a better design for the public displays brief, as it allowed my to test and improve my work numerous times.

The first iteration included the examination of the environment and considering the given problem and requirements. In order to examine the space, we created “Independant Dorset” posters, according to what we believed would appeal to the target audience, and work well in the space. Our aim was to create a poster that would encourage the engagement of the audience, in order to understand what the audience in the space were participating in, and how to grab their attention. In order to create this poster itself, we used the iterative design process; initially we started by understanding Independent Dorset campaign, and how we could communicate the message in poster form, we then proceeded to brainstorming rough initial ideas, from which we gathered feedback and chose to develop the most effective idea as a prototype. After testing our initial idea as a prototype we discovered a lot of information about the audience of the space, and how they interacted with the space.

The information I gathered on how the audience interacted with the space, it became clear that the use of posters was not very effective.  It was obvious that there was a large flow of staff and students in the area at either 10-5 to or 10-5 past every hour, when people were rushing to get to their destinations, however at times in between, and after 5pm, the area is quite empty. We noticed that the people going through the area would not pay attention to their surroundings at all, however people waiting in the queue for Costa or sitting in the chairs were almost always in groups, and were preoccupied in conversation with each other.

We also noticed the surrounding itself was extremely busy, with already a lot of posters and activity on the walls. The walls themselves were coloured orange and white, so next time posters created could be made to contrast the colour of the walls. The posters on the wall were very brightly coloured, and due to the extensive amount, very distracting.  Our main findings showed that the space was very fast moving and heavily text based installations to not work very well. We also found the people using the space were not very interested in the visuals of the space, however did pay more attention to the digital screens.  For this reason it became clear that the use of the screens would be most effective, as they received more attention that the posters in the area.

I then began to experiment using processing, and researching previous interactive installations. It became clear to me that I would choose to create something simple, yet effective, as I found coming to grasps with the processing language quite difficult. The concept I chose was based on the social constructing of identities using social networking websites, such as Facebook. For this reason I chose to use the facebook default photo as it is easily recognisable, and will encourage engagement.

I stumbled across an existing installation where a persons image was replaced by a block coloured silhouette of their body, this influenced my idea to use Blob Detection, in order to create a blob of the persons body int eh same colours of the Facebook default photo, however after testing my Blob Detection prototype I found that it was not very effective, as instead of tracing the image, the code traced the light in the area. This created random blobs on the screen, which did not apply to my idea, as it did not create the image of the default picture like I desired.

As a consequence to Blob Detection not working, I proceeded to use Face Detection, as I found that it recognized the faces more accurately.  I chose to create an installation where a persons image would be replaced by the default Facebook photo, instead of recreating the photo itself. 

After testing the Face Detection on some classmates I found that my initial code did not fit the screen well, and also did not mirror the video feed, however flipped the image.  After this was improved I retested in the Foyer, and found that the code worked reasonable well. I found that as people walked by they paid attention to the installation, due to the fact that they saw a familiar image.

Based on my user testing I feel I was definitely able to achieve my goal of creating a familiar image that the audience could relate to and engage with. The use of a camera based installation made people passing either stop, or slow down at least, as they noticed that the image was mirroring them as they passed by. Although it wasn’t a great response, the fact that a number of people altered their usual performance in such a busy environment showed some interest.

My idea was intended to reflect the concept of online identity construction, and the unreliability of forming relationships online. The intention of replacing a persons face with Facebooks default profile picture, was to communicate the idea that self-representations online may be deceiving. I researched htese ocncepts by looking at the MTV show Catfish, where the misuse of online personas are exposed, highlighting the flaws of constructing identities online. I also looked at the work of Benjamin Gaulon, where the blurring of boundaries of online and offline are explored, and the issue of privacy online. The use of a camera based installation in such a busy space, like the foyer, challenges the notions of privacy itself, and highlights how the increase of digital media platforms is blurring the public and private spheres.

If I was to repeat this process I would definitely choose to display my work at a busier time, therefore receiving a larger amount of engagement. Researching the space led to the conclusion that the busiest times were between lessons, as people passed by to get ot their destinations, therefore I would like to display my work at an earlier time, and leave it over a few days in order to record the reactions received over time. Another aspect I would improve is the use of the whole body, as using Face Detection allowed me to only detect the face, and create an image over the face of the viewer. I would like to experiment with different libraries in order to create a full silhouette as I originally intended. Overall I was pleased with the outcome of this project, as I managed to successfully create an interactive installation, which did produce a small reaction of the audience.