A challenge we will come up against is getting everything at the right scale. How large should the world map be relative to the size of each photo? Should photos that are more important be larger? How will we get an accurate scale relation between different models (for example, the La Perouse area model and landmark models such as the Museum and Tower)?
If one tends to stay put while wearing an Autograher, the resulting 3D spatial representation of the images gets much harder to view and interpret. If we’re going to make some kind of narrative, it will either have to be continuously moving or we will have to find a way to expand the close together data
We were split up into two groups for the project so that we could focus on different approaches to the Project. I am to be more focused on the technical aspects of the project which is a relief because I’m terrible at all this critical analysis kind of stuff.
While discussing how we might come up with a narrative arc, Hannah made the important point that presenting a photo trail narrative is too simplistic. We should be thinking about the interaction viewers can have with the work and the way they can explore the concepts of memory loss whilst taking part in the experience. Maybe we can allow users to explore the landscape, coming across fragments and unclear snapshots and attempting to salvage the rest of the memory.
Some notes I took during the class (Amanda Barnier was there):
- Why La Perouse was picked as the location
- Not as touristy as somewhere like Sydney Harbour
- Make use of the features of the landscape
- Engaging a bodily experience
- Physical actions, beginning and end points
- Encoding and layering of memory
- Identity memory
- Sensory memory and SenseCam, SenseCam retains everything we usually discard
- Shared experience of childbirth, shared but dissimilar
- Being behind your own eyes or being off and intellectual, rumination.
- Oscar Pistorius and the cross-examination memory reconstruction alteration
- The flood of sensory memory captured by a SenseCam
- Ideas for the Project
- We could follow couples around the area as the explore the area
- The water, digging things up, time capsule. Time frames
- Views from a lower angle indicates a child’s perspective
- The dimension of time
- Different people observing the same things can remember different things. We encode memories that are relevant to us. SenseCam grabs everything.
- Shells and shell arrangement from La Perouse history
Tanya and I were the only people that came to the mapping field trip. We drove to the proposed location where the project’s narrative will be taking place. Volker brought some form of DJI Phantom drone that holds a GoPro pointing directly at the ground.
The idea is to capture a whole bunch of photos from an aerial perspective and use software to stitch them together to create a huge 3D model of the area. I had wondered if it would be easier for the software if the camera was angled and was able to more easily capture a larger surface area, as well as the sides of any structures on the landscape. It didn’t really matter that much because it was hard to get the drone further than this off the ground. The drone is expensive and it was not handling well with all the wind near the ocean so it was dangerous to take it any higher (we were envisioning a light wind picking up and depositing the drone in the bay).
All was not lost, instead we took a whole bunch of photos of the landmarks nearby. We then used the 123D Catch app from Autodesk to do all the hard work analysing and converting the photos to a 3D model. I’m quite surprised how well it turned out.
The software promises that all you have to do is upload two orbits (from different angles) of the object you want to model and it will create a 3D model. I only did one orbit of each object (less than one in the case of the museum) and it still turned out pretty good. Not great, but it’s definitely something to work with. Besides, a faithful recreation would be boring, it’s much more interesting to look at a melting, deformed and grotesque building.
The one that turned out the best is the tower, probably because it is a simple shape. I was impressed that the software was able to pick out the indentation of the windows from just my photos. I think if I did another orbit, maybe with the camera on a stick, the model would turn out with even better quality. It might also get rid of the strange artefacts at the sky at the top. It is possible to manually remove the sky but having used the software I now have a better idea of how what to capture.
I’m glad COFA decided to take a week off this week, because CSE certainly didn’t. I spent the week frantically doing computing assignments and not much else. Jack worked up the courage (I wouldn’t have) to ask if we could do a more technical assignment and it looks like it’s all going ahead.
Pros: we’ll do stuff like messing around in Unity to deal with projection on the special immersive surface and other cool technical things (this is pretty much the reason I signed up to this project).
Cons: we have to meet up during the holiday next week.
Oh well, I have a feeling it’s going to be worth it. I would love to work with more people from iCinema. I’m looking forward to the drone exercise too, sounds fun.
Generally my presentations are pretty bad if I haven’t rehearsed enough or if I don’t know what I’m talking about. This was one of those occasions. I could tell a few minutes in that my brain was not functioning properly and I was droning on about R’s and remembering and lifelogging while skipping important points and not really saying anything of importance. Luckily I switched tone halfway through and managed to veer on to the topic of Google taking over the world, which stimulated discussion because everyone has a point of view about that.
I discovered in my research that Google’s entire business revolves around logging everything you do and helping you to ‘Remember intention.’ They want to provide all the information you need just before you need it. Pretty handy but most people are afraid of the ramifications of Google’s rampant data collection (and you should be too, with a healthy dose of skepticism and research).
There were other presentations, but my brain was cooling off so I didn’t manage to absorb much from them. I remember there being wine though. And not many people turning up.
Some key points from the presentation:
There are the Five R’s of human memory.
Recollecting is to do with the location of lost objects, faces, names and specifics like that. An Autographer or SenseCam can help with that. Reminiscing is a specific form of recollecting and is done for sentimental purposes. Facebook is good for this, as is the share functions on most things. Retrieving is for getting specifics like documents or other information. Cloud storage and other business tools are examples of tools that deal with retrieving. These tools are not visually oriented, they are primarily utilitarian. Reflecting is used to discover useful information from patterns in past information.It can be for learning and self-identity. Remembering intention is the final R, its deals with prospective events and planning the future. It is not relevant to the objectives of most lifeloggers, although it seems to be the way online services such as Google are heading.
I wasn’t paying too much attention to this lecture as I was trying to get a boat load of other assignments done for the week. We discussed the project and where the location for shooting might take place.
We watched Memento, which I happened to watch at the start of the course before we had any classes. It’s helped me to understand the how that particular type of amnesia could affect a person, I think the concepts presented by John Sutton and Amanda Barnier were easier to understand having seen the film.
We were also reminded that the assessment task would take place next week. Unfortunately I haven’t yet prepared for it…
Amanda J. Barnier
Experimental Cognitive Psychologist
7 things about memory
- Memory is everything
- There are past memories and
- it is important to remember things that are going to happen
- knowledge but not memory (“remember” is in quotes)
- Semantic dementia is semantic memory loss e.g. can’t remember what a dog is
- cultural memories (memories of the holocaust or Abraham Lincoln)
- What is memory for?
- Learning language, knowledge of the world, to plan the future
- Serving the self
- who we are, our connection to past, or a database of self
- we can’t remember everything
- To promote relationships (social)
- teaching, developing and maintaining intimacy
- A balance in memory:
- remembering has priority in our lives ( we worry: do I have a good memory?)
- forgetting promotes cognitive efficiency
- intrusive memory is when we can’t forget
- forgetting is not a failure
- How memory works:
- processes (like a computer):
- external event is placed in sensory memory
- given some attention this is then placed in short term memory
- with rehearsal this can be put in long term memory
- recall comes out of short term memory
- memory is the product of these sequence
- The essence of remembering:
- memories are a record of experience with reality
- memories are samples of experience
- remembering is a constructive mental process
- memories are part of the present moment
- memories are inherently subjective
- and can only be checked with in-corroborating evidence
- Influences on memory:
- time – memories become increasingly inaccessible
- reminiscence bump
- discussion with others
- police don’t let witnesses talk to each other because they can egg each other on.
- Judging memories
- Claire has both types of amnesia, retrograde and the other one (?)
- Available memory (stored and could influence you but may not) and accessible memory (things you can get right now)
- Listening to lectures doesn’t work! You have to be concentrating for it to enter your memory
- We don’t remember alone, we remember with others
- Infantile amnesia – the bit of life you can’t remember when you were first born
- if you show an amnesia sufferer someone else’s Sensecam pictures, would they have false memories?
- Reminiscence occurs only if the memory affected the fabric of your daily life
- Hotspots in trauma memory
- Remembering from the outside
- I remember past events from my own point of view or i remember from the observer perspective
- Remembering from outside must be a fabrication
- Can observer memories really be memories? What is the significance of changing perspectives
- How do I figure in my memories?
- Movement skills (sport, dancing): sometimes it helps to visualise yourself
- Embodied or kinaesthetic perspectives / cognitive and affective perspectives
- blending the above views together, fuse into interesting combinations
- Perspective in remembering, dreaming, navigating in spatial cognition, own body representation and gesturing
- Does truth in memory require preservation and reproduction?
- Field perspective memory can be constructed
- Change in memory recounting is the norm, it doesn’t mean you aren’t telling the truth
- Reports from your own eyes include more fantasy
- Embodied remembering
- Distributed cognitive ecologies: using other people to remember
- I do not identically reproduce a tennis stroke every time, it is a manufactured movement according to some schemata
- Athletes etc.: I need to stop thinking and just do
- There may be bodily skill.
- The body moving in place is not separate from any form of memory
- Research on out of body experiences is now the most prestigious cognitive science now (Curious).
- epilepsy sufferers have more out of body experience
- Women have more external perspectives of themselves due to the male gaze
- Wayne Rooney visualises himself from the outside when playing sport
- Story of the woman who designed her house the same as her childhood house to gain approval from her father
I was given an Autographer this week to try out. Here are some impressions.
The GPS lock took forever to acquire. Actually in a couple of cases it just didn’t pick it up at all and there is no location data attached to my photos. I was standing still out in the open for about 10 minutes before I decided to give up and just walk around without the lock. When in the settings screen, you can see how close it is to gaining a lock. I was under the impression that if it said ‘Low’ then that means it has a lock but it is inaccurate but it turns out this is a measure of how close you are to gaining a lock. You have to wait for it to go through Low, Medium, High and Lock before it will collect GPS data. Another time I gained a lock within 3 minutes so it appears to be very flaky.
The build quality of the device is pretty crap, it feels about the same quality as a $30 MP3 player. The buttons rattle around and the camera shutter is loose and clicks like a cheap toy. It’s unfortunate for a device that is around $400, I’m glad I didn’t pay for it. I’m not sure how they justify the price when there are phones with all the sensors included in the Autographer and more for around the same price.
The image sensor is not great, pictures are often blurry, have blown out highlights, and low light quality isn’t too good. It also auto-exposes to strange results.
Thoughts I had whilst wearing the device:
- Don’t wear headphones, they get in the way.
- We’ll be capturing a lot of photographic information and their locations, including peoples faces, car number plates and that kind of thing. Will we need to be mindful of privacy concerns. Will we have to blur out this information, and is there an easy way to do this?
- When I’m walking somewhere at my own pace, I often turn my head and glance around at my surroundings. I often notice little things happening around my, like interactions between people. This won’t be picked up by the camera if I’m turning my head to the side too much.
- I felt pretty self conscious when I was wearing the device, I thought people would be weirded out by being recorded. Most people didn’t seem to notice, and no one asked me about it. I wondered if there was an algorithm we could run on the photos to detect if someone was looking at the camera.
- It was looking like it was going to rain for most of the time I was using the device, I would like to know how water resistant it is.
- I read something about taking photos to remember something actually makes you forget it more easily. So if I’m wearing the Autographer, will I more easily forget what happened during the day? Do you more easily forget lectures when you write down notes?
- I often notice people’s mannerisms and actions, but will I remember them when I look back over the photos?
- If we stuck two Autographers together, would they take photos at the same time since they run with the same algorithm?
- A few things happened to me that couldn’t have been picked up, like a bus splashing water on the back of my legs.
- I pressed the sequence button a couple of times, and sometimes I couldn’t figure out how to stop them. It’s difficult to see the screen and operate the device when wearing it on your shirt collar.
The Autographer has very limited functionality. It provides interfaces to create videos of photo sequences, but thats the only way you can export anything from the program. Data takes a back seat to the photos, all the data is there in the corner, and there is a map to show where each photo took place, but there doesn’t seem to be a way to export the data without diving in to the Autographer library files. Even then it appears to have changed the format of the data from what was stored on the device. Volker has said we should copy all data from the device itself instead of using the software.