Seamus Nash: DECO3850 Portfolio

Hey everyone! Welcome to my Portfolio for DECO3850. Click on any of the links above to start looking into my Power Pose Elevator Project! A little about me, I am a final year Information Technology student majoring in Software Design. My interests include going to the gym, playing video games and hanging out with friends!
Below are some links to some other tasks conducted in the course over the semester. Click on them below to view them.
Prototype Report
Prototype Video
Original Team Video
Team Proposal Report
user testing results

"The Ideal"

The final product was developed while discovering what would happen when trying to enhance a mundane space. Our team came up with the idea of trying to enhance the mundane space of an elevator. For the team's perspective, we wanted to be able to develop something that enhanced the experience of an elevator ride by matching poses and or charades through voice, and movement. We first wanted to do a dancing elevator but when we received feedback from this, it reuslted in us changing it due to the ride being a quick experience and a dancing elevator wasn't feasible. From here, we brainstormed and decided to do with movement, charades and voice. Each member created a prototype that focused on a particular perspective which was a different audience for each prototype.
For my individual perspective I tried to enhance this space by creating a pose matching elevator that focused on the target audience of office workers. The intended experience for this product is represented in the GIF below shows the intended interaction with someone trying to imitate the Wonder Woman Pose. The image below would be how it could be installed in an elevator as a screen with a pose to imitate. I chose this area as office workers use an elevator daily and it would be interesting to see what they could do. See design process to see how I decided to come up with the power pose!
Image from: CrownTV

"The Actual"

The video below covers the intended experience in a live fashion. This actual scenario does show a whole body pose as this was a limitation that could not be worked around. It also represents a shed as an elevator and uses printed out elevator buttons to simulate pressing a level. Once a level is pressed, a power pose will be displayed and the user has 3 seconds to hold the pose, as this is the average time to give enough time to pose and exit. Once finished, the user will be prompted with a correct or incorrect response.

Technical and Physical Makeup


The images to the left show the physical prototype from its beginning to end function. As you can see, the elevator was constructed by using my shed as a simulated elevator and a piece of paper of elevator buttons to simulate an elevator ride.


This project has been mainly developed using Posenet, a vision model in JavaScript that can be used to estimate the pose of a person in an image or video by estimating where key body joints are. These key points are specific such as the left eye, nose, left knee etc. It also gives a pose confidence score and a keypoint confidence score, which determines the overall confidence in the estimation of a pose or a keypoint. It can be used to hide poses/key points that are not deemed strong enough. It also prints out the position of the key points as an x,y coordinate of the pose and image for each.
Image from: Medium
To actually match poses, a technique called cosine similarity was used. This is a measure of similarity between two vectors. So it measures the keypoints of the pose image shown and the keypoints of the person/s bodies. The code below shows how each keypoint is put into their array then subsequently compared. If the comparison is smaller than 1 then it is assumed that a match has been made. If not, then the pose is incorrect.

The Design Process

This project went through a rigourous design process, below is a diagram representing the minimalistic process taken throughout the entire project and highlights the user engagement made and thus, reveals the impact of decisions made in process.
Color key:
Light Blue Requirements
Orange Designing Alternatives
Lightgreen Prototyping
Pink Evaluating

Amy Cuddy - TED Talk on Power Poses. Image from: TEDBlog

First iteration on posing

Through this process at each bubble, an iterative process was done so then each stage and change made was verified by the requirements. Some user and background research was completed to see what these requirements may be. From the interviews with users, it became apparent that an idea would be to hype people up before meetings/interviews. From here I decided to commence in creating something that does this "hype". When looking into a TED Talk on Power poses from Amy Cuddy, this became apparent to use. This was because she argues that power-posing can be more effective than traditional confidence-boosting exercises, like telling yourself how great you are. She describes power poses as expansive and open. From here, after performing the first iteration on which the users were able to just power pose with one person, it became apparent that their were some issues. As you can see in the diagram, the user wanted some more signifier to show what to do and also wanted to do it with someone else. Audio was added to rectify this and the final product was changed so that multiple people could use it.

Reflection on Outcomes

The desired outcomes for this project were to try and make an elevator an enhanced experience. This project relates to the other work shown in the domain as it is able to enhance a mundane space and is a physical, playful and interactive product that is fun as opposed to a mundane one. This is done by getting user to copy a power pose shown on the screen. Also, it was essential that the project was able to enstill confidence in office workers. This has been demonstrated by placing the pose matching area in an elevator and getting users to perform power poses as these are able to enstill confidence.
In terms of the final product's desired outcomes, the final product meets most of the outcomes but the only outcome that wasn't able to be met was the ability for the elevator to be more portable and furthermore, placed in an actual elevator. If this was completed, modifications may need to be made to the current product to suit this setting. Due to a worldwide pandemic and limited resources, their are further iterations that need to be conducted. Also after the last round of user testing was conducted, it was found that users wanted more than one power pose to copy as they wanted more variety and due to time constraints this wasn't completed. The user is able to experience the entire posing sequence and is able to exhibit confidence and more positivity. This was measured by investigating facial expressions and general confidence which was all done by observation and interview questions. In terms of human values, users were able to accept the physical layout of the elevator as it didn't simulate the ideal product. Finally, I feel that this semester has been a wild ride, but I am glad with the result I have come up with even in a pandemic.

Team CDI

Below are the links to my fantastic team members who were also investigating this space. Click on their images below to view their portfolios

Dimitri Filippakis: Alistair Harris: Anson Cheung:
User Experience Design and Information Systems Major Finance and User Experience Major User Experience Design Major