October 2016 – Page 6 – Exploring Digital Heritage

A trend has begun within these classes moving towards this class today I think. Ways in which we can utilise cultural heritage data to make it more engaging, and accessible using digital tools – and today really empathised that with the use of computer vision technology. We have looked at neatening data, making it into a variety of visual graphs and so forth. This week was all about images, and how to help computers see images in ways that are more useful and tangible for us (and sometimes accidental like Redaction ‘Art’). The most visceral and moving example was the images of the faces of people under the White Australian Policy. I think this is a beautiful way of fleshing out an issue that is widely known, but more in a way of dates and laws, rather than the actual who. Visual aids appeal to our humanity because a face is so recognisable and emotive, it instantly builds empathy because this is another person just like you or me.

Tim was discussing that the goal for another example, the Trovebot, putting faces on people who sent in photos was to try and connect people to the past in new ways. I love this as it continues to makes this data significant, because without interaction its meaning is lost. Face detection technology is a lot of fun too and I can see how this can be used in a variety of different ways, perhaps even at cultural institution with a photo booth like set up putting images of faces from the museum on to your own.

The face detection subject just reminded me a lot of the Snapchat filters that are popular, and sometimes they do have historically relevant ones (such as Marilyn Monroe on her birthday or the faceswap option where you can upload and image to swap with your own which I have seen is quite popular to do at classical art galleries). This is engaging, shareable and fun and this is another outlet that is potentially significant for cultural heritage as it is very good and well known technology.

Facial recognition on the other hand, a little creepy, but without obvious benefits used by the right people. Being able to identify historical figures in object would be great for example.

This is the face replacement with @facedepot.

It’s quite interesting and useful. This belongs to face detection, not recognition. Actually, face detection has already used for entertainment use, like Snapchat, you can create interesting video with interesting face in real time.

I’ve tried the Face++, it can not only detect human face, but also tells you the gender, age, races, and other information.

I think it can be used for department shopping industry, when people want to try some clothes, they just need to standing in front of camera, or upload their photo online. Then, according to their gender, age, races, body size… the system can give suggestions to them. However, it can also raise racism and gender equal problems. Nobody can say Asian people can’t wear what kind of clothes, or can. Nobody can say female can’t wear what kind of clothes, or can.

And for those hair styles that can blind facial recognition system, I don’t think it’s a big issue. Because with the development of technology, more advanced system will appear, and it won’t blind facial recognition system any more.

However, just like many people may think, the Aussie government’s new ‘weapon’ may raise privacy problems. So, I fully support government’s behaviour, but they need to make sure people’s privacy can be fully protected, only used for government departments. Not sold, not leaked, not like census.

And yes, facial recognition can be very helpful for heritage protection, instead of human eyes. Maybe you can find Emperor Qin Shihuang was walking in crowd at a ancient China painting?

For the last video at week 9 handbook, I think it can raise fraud issues. It looks like Snapchat, but if some people believed it, maybe someone use granddaughter’s face to fake grandmother of grandfather. And it’s not only photo, it’s video, with the voice can also be imitated with latest technology.

I have tried IBM demo page, and the result is right.

Month: October 2016

Week Nine: Face It

Week 9 Reflection_Yizhuo Zhang u3125186