You are being invited to participate in a study entitled “LatentKeypointGAN: Controlling GANs via Latent Keypoints”.
The purpose of this study is to evaluate the utility of LatentKeypointGAN: Controlling GANs via Latent Keypoints. This is important because it allows us to understand whether the proposed tool is capable of generating high quality images and disentangling images in parts. We will show an example of part disentanglement in the example questions.
If you volunteer to participate in this research, your participation may include:
• Filling out a background questionnaire that asks about your experience with computer technology and data analysis applications, as well as basic demographics, such as age and gender.
• Completing computer-based tasks that involve comparing portrait videos and other videos of persons, objects and numbers.
• Filling out a questionnaire about the computer-based tasks and tools you experienced.
The session is expected to take approximately 30 minutes and, depending on the research stage, will take place at one of the following locations:
• Your place of work, including the home office.
• Remotely via a web interface and introduction by phone or a teleconference application such as Skype (without video).
There are no known or anticipated risks to you by participating in this research.
Taking part in this study is entirely up to you. You have the right to refuse to participate in this study. If you decide to take part, you may choose to pull out of the study at any time without giving a reason and without any negative impact. If you do withdraw, we will ask whether we may use your data for data analysis and destroy your data otherwise. Moreover, you will be asked to choose whether your participation in this study will be attributed (for example, by an explicit acknowledgement by name for your contributions towards the work in a publication) or confidential. If you choose confidentiality, your confidentiality and the confidentiality of the data will be protected by identifying data only with a participant number rather than your name. All data stored in computer files will be password- protected. We will not audio/video-tape your session. Only the principal investigator and co-investigators will have access to any personal information you provide.
We will give access to the study data (anonymized) via a research repository, earliest three months after this study. From that point on, participants will not be able to withdraw their data. Since all personal identifiers are removed, this poses minimal risk. We anticipate that the results of this study will be shared with others in the following ways:
• Published articles
• Conference presentations
• Theses
• Internet project descriptions
Data from this study will be disposed after five years or earlier on request. Electronic data will be erased and paper copies will be shredded.
If the subsequent questionnaire is completed, it will be assumed that you have received a copy of this consent form and consent has been given. You can withdraw your participation at any time.
![]() |
![]() |
|
A | B |
Choose the image (A or B) with higher face quality (background is not considered). If you think A and B have the same image quality, please select "They are equal".
![]() |
![]() |
![]() |
![]() |
|
A1 | A2 | B1 | B2 |
Regardless of image quality, choose the pair (A or B) with editing eyes better preserves the remaining facial features (better locality). If you think A and B have the same quality of part disentanglement, please select "They are equal".
![]() |
![]() |
![]() |
![]() |
|
A1 | A2 | B1 | B2 |
Regardless of image quality, choose the pair (A or B) with better identity preservation (recognizing the same person) while changing expression. If you think A and B have the same quality of identity preservation, please select "They are equal".
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
|
A (appearance) | A (shape) | A (combined) | B (appearance) | B (shape) | B (combined) |
Regardless of image quality, choose the 'combined' image (A or B) that better preserves the outline of the face and its parts from their respective 'shape' source. If you think A and B have the same quality of shape preservation, please select "They are equal".