User:VanessaLattas
From REU@MU
Week of 6/12
MON 6/12
- Drafted literature review, annotated bibliography sections of final paper
- Investigated the available AI generative tools for text-to-video, text-to-photo, and text-to-speech
- Documented a list of these tools, their features, and their drawbacks as it relates to project goals
TUES 6/13
- Met with mentor to share progress and troubleshoot roadblocks
- Evaluated the storytelling capacity of AI by reviewing materials on ACM database, GitHub
- Compared animated with live video and their independent ability to allow connection with audience
WED 6/14
- Attended REU meeting and presentation
- Demoed around 5-10 text-to-speech applications, software tools, etc.
- Assessed the emotions/tones appropriate for mental health storytelling (frustration, anger, helplessness, etc.)
- Tested applications to determine which were most realistic depictions of such emotions/tones
THURS 6/15
- Attempted implementation of Phenaki Video in Pytorch
- Requested access to Meta's Make-a-Video
FRI 6/16
- Accessed DALL-E premium version
- Used generic prompts from given subthemes previously extracted from PPD subreddit in previous research
Week of 6/19
MON 6/19
- Began process of creating 2 draft videos, one still, one animated
- Edited scripts from Reddit blog(shortened, corrected grammatical errors): one with social factors theme, one with hardships
TUES 6/20
- Met with mentor to share progress and troubleshoot roadblocks
- Created an animated video using Steve AI tool
- Created live still video using Designs AI tool
WED 6/21
- Attended REU meeting and presentation
- Met with mentor to share progress and troubleshoot roadblocks
- Presented draft videos to mentor and implemented feedback, editing in Vimeo and VEED
THURS 6/22
- Used DALL-E and Deep AI as primary text-to-image tools to prepare 5-10 templates for each of 4 themes
- Began generating templates for subthemes as well
FRI 6/23
- Prepared mini presentation on progress made thus far